agents.md holds the instructions, claude.md points at it

opencode discovers only AGENTS.md, Claude Code only CLAUDE.md — so with the
content living in CLAUDE.md, every doc here was invisible to opencode. Swaps
all five pairs (root, src/apps, src/databases, themes, plugins): AGENTS.md now
holds the full text verbatim, CLAUDE.md is a short pointer. docs-check's
permission-count regex now reads root AGENTS.md instead of CLAUDE.md; the
workspace-level officer-dev/CLAUDE.md check is untouched, out of scope.
This commit is contained in:
2026-08-27 06:35:31 +00:00
parent 185aa3168e
commit 2360a3ef60
11 changed files with 1280 additions and 1291 deletions
+552 -30
View File
@@ -1,41 +1,563 @@
# AGENTS.md # CLAUDE.md
Guidance for any coding agent working in the Officer platform repo. ## Project Overview
**The instructions live in [`CLAUDE.md`](CLAUDE.md). Read it first — this file only points there.** Officer is a self-hosted platform built around one person — the server owner — which since 2026-08-07
also admits **additional accounts holding a strict subset of it**. It bundles an AI agent, a terminal,
a file browser, a code editor, a bitcoin wallet, a remote desktop and customisable dashboards
behind a single web app.
Kept separate so agents that look for `AGENTS.md` by convention find the same guidance as those that **The owner/member split, and where the line falls.** This file said "single-user is a hard invariant,
look for `CLAUDE.md`, without the two drifting apart. not a stage" until 2026-08-07. That is no longer true and had already stopped being true when it was
written: `users` holds six rows. The accurate statement is narrower and more useful —
## Orientation - **One owner.** User id 1, role `Super Admin`, created by `POST /auth/bootstrap` while the table is
empty, pinned there by a CHECK constraint. The owner bypasses every permission check.
- **Other accounts get only what their ROLE is granted.** Roles are `Admin`, `Member`, `Developer`;
grants live in `role_permissions`, keyed on role, never on user. Absence denies — there is no row
meaning "no", so an empty table is a server where members reach nothing but their own profile.
- **Some things can never be shared, structurally.** Tasks, items, desktop and browser are
`kind: 'execution'`: they run as the owner's OS user in the owner's home, so there is no level of
"read" that makes them safe. They have no level at all and the grants API refuses to store one.
- **And some are shared only because the kernel enforces it.** Terminal, chat and files are
`kind: 'confined'`, added 2026-08-11 with per-user Linux accounts. They still touch the filesystem
and still run processes — but not the _owner's_, because the account has its own Linux user, its own
home, and the kernel refusing everything above it.
Officer is a self-hosted platform built around **one owner** (user id 1, role `Super Admin`, who The distinction earns its keep in one place: **a confined grant means nothing without that Linux
bypasses every permission check), which since 2026-08-07 also admits **additional accounts holding a user.** `authorize.ts` drops it for an account whose `osUser` is null, so "granted but unconfined"
strict subset of it**. Roles are `Admin` / `Member` / `Developer`; what each may reach is decided by resolves to no access rather than to the owner's home — which is what it would otherwise resolve to,
per-role permission grants, resolved on every request. since `getOwnerHomeDir` ignores the email it is passed. That rule lives there once and covers the
HTTP routes, the websocket doors and the dock together.
If a design question turns on "which user", the answer depends on the surface: real for the **app** So "which user is this" has a real answer for the **app** surface (gitea, music, photos,
permissions (gitea, music, photos, calendar…), and still always **the owner** for anything calendar…) and for the **confined** one (terminal, chat, files), and is still always "the owner" for
that executes code or touches the disk — terminal, chat, tasks, files, desktop, browser are anything under `execution`.
`kind: 'execution'` and can never be granted. `src/servers/permissions/registry.ts` is the authority.
**Mounting a router without a registry entry makes the server refuse to boot.** Read the "Permissions" `src/servers/permissions/registry.ts` is the authority and reads as the design document for this.
section of `CLAUDE.md` before adding one. **Mounting a router without a registry entry makes the server refuse to boot** — see "Permissions"
below before adding one.
This file previously described Officer as strictly single-user with "no tenancy, no roles, no user **Account creation exists.** `POST /api/users` (owner-gated), with a form under Settings → User Management:
management". That was written to correct an _older_ drift in the opposite direction — a fictional it validates the password with the same rules bootstrap uses, refuses a role that is not a row in `roles`,
multi-user intranet with a user-invitation API — and it overshot. Both are now superseded by the refuses `Super Admin` outright because there is exactly one owner, optionally takes an inbound SSH key, and
paragraph above; treat the permission registry as the source of truth over either. provisions the Linux account. `POST /users/:id/provision-linux` retries that half if the host was not ready.
This repo is one of two. The other, `permissions/`, holds the agent's tasks, tools and skills as This file said "there is no way to create a second account" for some days after `69a31051` shipped one, and
plain files, and is where most changes belong — adding or changing a task needs no code change here `docs/roles-and-teams.md` sequenced a whole phase around building it. Neither is checkable by
and no restart. `bun run docs:check` — it compares numbers, and this was prose. Worth knowing as the shape of the failure:
the counts stay right and the sentences rot.
- [`CLAUDE.md`](CLAUDE.md) — this repo's architecture, conventions and code style ## Architecture
- [`docs/working-on-officer.md`](docs/working-on-officer.md) — the deployment-wide guide: which
directory a change belongs in, how to run and verify it, the task system, and the failure modes
worth knowing about
- `CONVENTIONS.md` — component organisation and React patterns, with rationale
- `TODO.md` — current direction; takes precedence where it disagrees with anything else
Treat the code as the source of truth where anything disagrees with it. One Bun process (`src/server.tsx`) serves everything:
- the React SPA, via Bun's HTML import of `src/apps/officer-web/index.html` (HMR in dev)
- the REST API, a Hono app mounted at `/api` (`src/servers/hono.ts`)
- eight WebSocket providers — terminal, chat, task-runner, pipeline, cliamp, cliamp-audio, desktop,
vault — plus a sidecar registration socket. `terminal` is a byte relay onto the pty sidecar's own
listener, not a translating bridge; `vault` is the same shape onto Vaultwarden's notifications hub.
- ~~a browser relay on its own port~~ — switched off 2026-08-13, awaiting extraction into a plugin.
The extension and `api/browser/` stay on disk; the listener and the `/api/browser` mount do not.
Long-running and privileged work lives in **sidecars**: separate processes that dial back in over
`/api/sidecar/register` and are tracked in `src/servers/sidecar-registry.ts`. PM2 runs them
(the generated `ecosystem.config.cjs` — see below): `officer` (the server), `officer-anthropic-proxy`, `officer-claude-code`,
`officer-opencode`, `officer-pty`, `officer-vnc`, `officer-music`, `officer-vault`,
`officer-slskd`, `officer-headscale`, `officer-transmission`, `officer-invoiceshelf`, `officer-wallet`,
`officer-photos`, `officer-notify`, `officer-caldav`, `officer-memos`, `officer-jellyfin`, `officer-gitea`
— twenty as of 2026-08-06, and a list that goes stale every time a sidecar lands. `pm2 jlist` is the
source of truth.
**`officer-anthropic-proxy` and `officer-claude-code` are not the same thing.** (The second was
called `officer-agent` until 2026-08-13; older docs use that name.) The proxy holds the Anthropic
credential and forwards API traffic; the agent is the process that spawns `claude`. They were one entry
named `officer-claude` until the sidecar-isolation work — which is exactly how the false claim that
"restarting officer doesn't disturb the agent" survived so long. Every sidecar is a PM2 peer of
`officer`, so **no sidecar is a child of the server and restarting the server does not kill one.**
Agents run **unsandboxed as the server owner**, with `--dangerously-skip-permissions`. This is
deliberate — it is the owner's own machine. Do not add a jail without being asked.
## Repository Layout
```
src/
├── server.tsx # the single entrypoint: routes, WS upgrades, static files
├── apps/
│ └── officer-web/ # the SPA shell — Screens/Authentication + Screens/Dashboard
├── servers/
│ ├── hono.ts # router composition; everything under /api
│ ├── _middlewares/ # auth, body parsing, the permission gate, rate limiting
│ ├── api/<feature>/ # one folder per feature, each exporting a router
│ ├── channels/ # send-claude-code / send-opencode — how /chat drives an agent turn
│ ├── queue/ # background job engine
│ └── sidecar/ # sidecar implementations + the wire protocol
├── databases/officer_db/ # the only database (Postgres + Drizzle)
├── extensions/browser-relay/
└── workspaces/ # shared packages, each a bun workspace
```
`src/workspaces/officerdev` is the biggest of these: the windowed "apps" (FileBrowser, Chat,
Terminal, CodeEditor, Desktop, Dashboards, Wallet…) that the shell hosts, behind an `AppRegistry`.
`src/apps/officer-web` is only the shell — screens, routing and settings.
**Path aliases** (`tsconfig.json`): `@/*` → `src/apps/officer-web`, `@/components/*` →
`src/workspaces/components`, `@@/*` → `src/servers`, `@/public/*` → `public`. Workspace packages are
imported by their package name (`officerdev`, `hooks`, `state`, `types`, `helpers`, …).
## Tech Stack
- **Runtime**: Bun (Node 22 or newer is enforced by a `preinstall` check)
That check demanded _exactly_ 22 until 2026-08-12. The reason was a `node-pty` build
failure some months earlier, whose details were not recorded. It was relaxed to `>= 22`
after confirming node-pty ships **no Linux prebuilds** — its install script always falls
through to `node-gyp rebuild`, so it compiles against whatever Node is present and there
is no ABI to mismatch. Untested on 24 at the time of the change. If `bun install` fails
building node-pty, or `officer-pty` cannot load its native module, restore the exact pin
first. The source build also needs `build-essential` and `python3`.
**Bun itself is held at the 1.3 series, on purpose and temporarily.** 1.4.0 was published
2026-08-20; the owner asked on 2026-08-22 to stay on 1.3 for a few months until 1.4 has been
watched. Two places carry that number and they move together: `BUN_SERIES` in
`scripts/setup/machine-setup/lib/dev.sh`, which resolves the newest `1.3.x` from the npm registry
and passes it to the installer as a release tag, and `engines.bun` in `package.json`, which is
what `bun run doctor` reads. bun ignores `engines` (verified — an install declaring `1.9.x` ran
clean under 1.3.14), so that field is the declaration and doctor is the enforcement.
**Do not run `bun upgrade` on a machine running Officer.** It takes no version argument — latest
stable or `--canary`, nothing else — so it jumps straight past the pin and says nothing. That is
the whole reason `doctor` checks the running version at all; `machine-setup.sh --only 'JavaScript
runtimes'` offers to pin a drifted machine back down. Lifting the hold is `todo/infra/`.
- **Language**: TypeScript, strict. `bunx tsgo` is clean — keep it that way.
- **Frontend**: React 19, React Router 7, React Query, Tailwind 4, shadcn/ui + custom components
- **Backend**: Hono
- **Database**: Postgres via Drizzle — `officer` for the platform, plus `officer_marketplace` on the
host that serves marketplace.officer.dev (see `docs/marketplace.md`)
- **Agents**: Claude Code and opencode, driven through sidecars; tools exposed over MCP
## Data
Two stores, and the split matters:
**Postgres** (`src/databases/officer_db`) holds the account, passkeys, settings, dashboards,
queue and pipeline jobs. Since 2026-08-16 there is a SECOND database,
`officer_marketplace` (`src/databases/marketplace_db`), and the split is deliberate: it is a public
catalogue served by a process with no login, no session and no user, so sharing a connection would put
an unauthenticated surface one query away from every private table. It runs only on the host serving
marketplace.officer.dev — `docs/marketplace.md`. One directory per feature holding `schema.ts` and
`queries.ts` beside each other; `src/schema.ts` is what `db:push` reads, and it lists the core tables
with the plugin ones commented out. Types inferred from the schema in `src/types.ts`.
**The filesystem** holds everything the agent authors. `OFFICER_ITEMS_DIR` (`$OFFICER_ROOT/capabilities`)
contains one directory per item under `skills/`, `tools/`, `tasks/`, `processes/`, `extensions/` — no
database rows, no scope tiers. `DATA_PATH/<email>/` holds the managed home and attachments. It also holds
`email_accounts/`, which is the **email plugin's** — the platform pre-creates the directory and never opens
what is inside it.
**None of those paths is configured.** Since 2026-08-13 `src/servers/data-path.ts` derives the install
root as `resolve(process.cwd(), '..')` and hangs `data/`, `permissions/` and `dockers/` off it. That
replaced `DATA_PATH`, `OFFICER_ITEMS_DIR` and `HOME_DIR` in `.env` — three values that had to agree with
each other and with the tree on disk. `assertInstallLayout` refuses to boot when the working directory
is not the repo, because otherwise a wrong `cwd` relocates the whole install silently rather than
failing.
Note `getHomeDir` (the managed home under `DATA_PATH`, now used only for NON-owner accounts and by
pipeline-executor) versus `getOwnerHomeDir` (the owner's real login home, where terminals, chats and
task runs execute — captured from `homedir()` once at module load, and it ignores the email it is
passed).
### Schema changes use `push`, not migrations
This database is kept in sync with `bun db:push`, which diffs the schema code against the live
database and alters it directly. **`drizzle-kit migrate` has never been run here** — there is no
`__drizzle_migrations` table. Change the schema, run `bun db:push`, done.
`bun db:gen` writes files to `officer_db/migrations/`, but nothing applies them; the old numbered
history was deleted because it had drifted from the real schema. Treat the schema code, not those
files, as the source of truth.
**Declare multi-column uniqueness as `uniqueIndex('uq_…').on(a, b)`, never `unique('uq_…').on(a, b)`** —
drizzle-kit mis-diffs named composite unique _constraints_ and re-creates them on every push, which used
to stop `db:push` on an unanswerable truncate prompt. Same for any foreign key whose generated name would
exceed Postgres's 63-character identifier limit: name it explicitly. See `src/databases/CLAUDE.md` →
"Composite keys" before adding either.
## Security Model
- `IS_DEV_BUILD` (`src/servers/build-env.ts`) is true **only** when `PUBLIC_BUILD_ENV` is explicitly
`dev`/`development`. Everything else, including unset, is hardened. Rate limiting and password
rules key off it — they fail closed.
- **There is no origin checking.** It was removed on 2026-08-13, along with `ALLOW_ANY_ORIGIN` and
`ALLOW_ANY_ORIGIN_MUSIC`. The flag defaulted to ON, so origin validation ran on no real install —
what came out was documented defence in depth that was already switched off. Origin was never
authentication here anyway: an app's `officer://<hex>` origin is chosen by the client, forgeable
outside a browser, and extractable from a shipped binary. The perimeter is the tailnet, and the lock
is a valid token on every protected route plus the permission gate below.
- JWTs are 30-day, blacklisted on signout, and invalidated by a password change (`passwordChangedAt`).
**The role is deliberately not a claim** — every authorization decision re-reads `users.role` from
Postgres, so a grant or a revoke takes effect on the next request rather than at next sign-in.
- A panic lockdown (`src/servers/api/auth/panic.ts`) is in-memory only and refuses every
authenticated request until the server restarts.
### Permissions — read this before mounting a router
Authorization is one system, and it is not in `userMiddleware` (which only answers "is this token
valid"). It is `_middlewares/permission-gate.ts` → `permissions/authorize.ts`, mounted globally in `hono.ts`
ahead of everything, and it re-verifies the token itself so it covers routes that never mount
`userMiddleware`.
- `permissions/registry.ts` — the single enumeration of what the platform can do, in five kinds:
`core` (every account, not deniable), `app` (**the grantable surface**), `confined` (grantable, but
only to an account that has a Linux user), `execution` and `admin` (owner only, and `execution` is
never grantable at any level). **21 entries as of 2026-08-27** — verified by `bun run docs:check`, which fails if this number drifts again: 6 core, 2 app, 4 confined, 3 execution, 6 admin (`todos`, the todo board, is the sixth). It was 22 until caldav became a plugin and took `calendar` with it — a plugin's permission is registered at runtime from its manifest, so it is real but not in this file's count. `notify` moved from `app` to `core` on 2026-08-24: it was grantable, nothing granted it, and every Member's app was silently 403'd registering its own push token. It said 27 for nine days. The execution count dropped by one when Minesweeper became a plugin — see `plugins/minesweeper/manifest.ts` for why that was a fix rather than a downgrade.
- `permissions/authorize.ts` — resolves "may this account do this". Owner short-circuits first; every
other answer is role grants plus core, with `execution`/`admin` stripped even if a row grants them,
and `confined` stripped for an account with no `osUser`.
**Every catch returns deny.** Grants are cached by role and the cache's whole invalidation contract
is `invalidateRoleGrants`, called by the one writer in `api/users/permissions-routes.ts`.
- `permissions/totality.ts` — `assertPermissionTotality` runs in `server.tsx` **before `serve()` and
throws**. Mount a router or a socket without a registry entry and `pm2 restart officer` fails,
naming what is missing. That is deliberate: the hole it closes was a Member 403'ing on
`GET /api/tasks` and opening `/api/tasks/pipeline/ws` with a 101 in the same minute, because Bun's
route table matches the socket before the `/api/*` catch-all that reaches Hono. A patch does not
survive the next door; refusing to boot does.
So **adding a router means adding one line to `CAPABILITIES`**. If the surface genuinely is not
user-gated, add it to `EXEMPT_API_PREFIXES` in `totality.ts` _with a reason_ — an unexplained exemption
is how the hole happened the first time.
The frontend hook `usePermissions` **fails open** on purpose: hiding a dock icon is a courtesy, the
403 is the lock, and an owner locked out by a transient network error is worse than a member clicking
into a refusal.
## Commands
```bash
bun dev # the whole app — SPA + API + WebSockets, watched
bun start # production
bunx tsgo # typecheck (not tsc)
bun test # tests
bun format # prettier over every dirty file — see the note below before running it
bun db:push # apply the schema to Postgres
bun docs:check # do the docs still match the code? REQUIRED before tagging a release
bun setup # runs scripts/install.sh — blank machine to running platform
```
**`bun docs:check` before every release, and leave it green.** It compares documented claims against values
derived at run time — the permission count, the PM2 entry count, every settings route Spotlight offers — and
exits non-zero on drift. A check must never compare a doc against a second hard-coded copy; that is one more
place to drift, and it drifts in the same edit that misses the doc.
It is narrow on purpose. It cannot read prose and judge it, only claims with a machine-checkable
counterpart — which is a small part of the documentation and the part that rots fastest, because a number is
exactly what nobody re-derives while editing the sentence around it. Both claims it found on the day it was
written had been wrong for over a week.
`scripts/install.sh` is only an orchestrator — it runs the two halves in order and does nothing itself:
`setup/machine-setup/machine-setup.sh` (28 sections: packages, tailnet, runtimes, docker, shell) then
`setup/officer-setup.sh` (11: pre-flight, layout, repository, dependencies, database, environment, secrets,
schema, build, services, verify). Either runs alone — `--machine-only`, `--officer-only`, or by path — because
a machine you already trust needs only the second. Both are re-runnable: each records the steps it finished
and skips them, so stopping halfway costs nothing. **Run it as yourself**; it re-execs through `sudo` when it
needs to, and on macOS never does, because Homebrew refuses to run as root.
Sidecar control is PM2, not npm scripts: `pm2 restart officer-<name>`, `pm2 logs officer-<name>`.
See `docs/working-on-officer.md` for which process a given change needs restarted.
### Installs are frozen. Never resolve a dependency you did not ask for.
`bunfig.toml` sets `[install] frozenLockfile = true`, so **`bun install` resolves from `bun.lock` and
nothing else** — it fails rather than quietly picking up a newer version, including a transitive one
nobody chose. Verified on bun 1.3.10: a lockfile that no longer satisfies `package.json` exits 1 with
`error: lockfile had changes, but lockfile is frozen`.
It is config rather than a habit because a supply-chain compromise does not wait for the one time
somebody forgets a flag. On **2026-08-04** eleven cache packages — `keyv`, `flat-cache`,
`file-entry-cache`, `cacheable-request`, `cache-manager`, the `@cacheable/*` scope, `ecto` — were
published with a `preinstall` dropper that harvested npm and GitHub tokens, AWS and Kubernetes
credentials, SSH and PEM keys, `.env` files and `.claude/settings.json`, then republished itself
through any npm token it found. It reached 434 further packages across 1,381 versions. This machine was
unaffected only because nothing had installed since 2026-08-02.
**To change a dependency:** edit `package.json`, allow the lockfile to change deliberately, **read the
diff**, and commit it. Note that `--no-frozen-lockfile` — which this file and `bunfig.toml` both used to
name — is NOT a flag bun 1.3.14 has; it is silently ignored and the install still fails frozen. Comment
out `frozenLockfile` in `bunfig.toml` for the one command, then put it back. The friction is the point — an unexplained lockfile change
in a diff is the signal this exists to produce.
**Do not** add `--no-frozen-lockfile` to a script, a Dockerfile or CI to make an error go away. The
error means the lockfile and `package.json` disagree, and that is worth a human look every time.
**Format your own files, not the whole dirty tree.** `bun format` globs `git diff --name-only HEAD`,
so it rewrites every uncommitted file — including work in progress that isn't yours, which then shows
up as unexplained whitespace churn in someone else's diff. Run `bunx prettier --write <paths>` on the
files you actually touched. `bun format` is only safe when the tree is otherwise clean.
## Code Style
- **Paradigm**: functional — pure functions, immutability, composition
- **TypeScript**: strict, no `any`. Type-only imports are required (`verbatimModuleSyntax`).
- **Comments**: minimal, and about _why_. Don't narrate what the code already says.
- **Async**: always async/await
- **Exports**: named only, no defaults
- **Files**: `PascalCase.tsx` for components, `kebab-case.ts` for everything else
- **Prettier**: single quotes in JS, double in JSX, semicolons, trailing commas, 120 cols
### Imports
Order: types → external → workspace → relative.
```ts
import type { User } from 'types';
import { useState } from 'react';
import { useQuery } from '@tanstack/react-query';
import { useClient } from 'hooks/useClient';
import { formatDate } from 'helpers/formatters';
import { useWebsites } from '../useWebsites';
```
In app code import types from `'types'`, which re-exports the database types. Only server and
database code imports from `officerdb/types` directly.
### Functions
Arrow functions for one-liners, regular functions for anything with real logic. No multiline
parameter lists — extract a params type instead.
```ts
type ProcessDataParams = { body: Record<string, unknown> | undefined; query: Record<string, string> };
export function processData({ body, query }: ProcessDataParams): Result { ... }
```
### Components and hooks
```tsx
type ExperimentCardProps = { experiment: Experiment; onSelect: (id: number) => void };
export const ExperimentCard = ({ experiment, onSelect }: ExperimentCardProps) => { ... };
```
Complex hooks return objects; simple state hooks return `as const` tuples. Event parameters are
always named `ev`, never `e`.
### API calls
`useClient()` returns typed verbs — `client.get<Foo>('/foo')` really does return `Foo`, so pass the
type parameter and let inference flow from it.
## Working With Me
- **Ask first** — confirm the approach before a significant change
- **Explore thoroughly** — read the related files before editing
- **Keep it simple** — no over-engineering, no premature abstraction
- **Be explicit** — no magic, no implicit behaviour
- **Stay focused** — note unrelated problems, don't fix them uninvited
- **Report honestly** — say what you verified and what you didn't
**Always commit and push when you finish implementing — including when it went wrong.** Don't wait to be
asked, and don't hold a branch back because it is unfinished, untested or turned out to be a dead end. The
history of the mistakes is worth having: a reverted commit and its message explain why an approach was
abandoned, which is exactly the thing that gets lost when a failed attempt is quietly discarded. Say what
state it is in — in the commit message and in `COMMS/` if another agent will pick it up — rather than
withholding the commit until it is good.
Commit messages: simple lowercase, no prefixes.
**Stage and commit in ONE step, naming the paths.** `git add <path>` then a bare `git commit` commits the
whole **index**, not the path just named — so anything a concurrent session had staged goes into your
commit, under your message and your session trailer. Use `git commit -- <paths>` instead, which does what
the two-step form only looks like it does.
This is not hypothetical and it is not rare: two agents share this working tree most nights. On
2026-08-23 commit `6e092c32` says "marketplace: minesweeper in the catalogue" and contains nine files of
another session's roles work, swept up mid-`git add`. The other session was right not to amend it —
rewriting shared history to fix a message is the worse trade — so the record was corrected with a
follow-up note instead. Cheap to avoid, awkward to undo.
Two things about the form itself, both learned the same night. `-m`/`-F` goes BEFORE the `--`, or git
reads the message as a pathspec. And it only works on **tracked** files — a new file must be `git add`ed
first, so a commit introducing one is `git add <new>` then `git commit -- <paths>`, naming every path.
Getting that wrong fails loudly, which is the good case: the release tag that followed it went onto the
wrong commit, and only because the tag was checked afterwards did that surface at all.
**An agent's commit names the session that made it.** A trailer, below `Co-Authored-By`:
```
Officer-Session: 656f5c85-e49a-4f1f-b9f3-042a04b4d043
```
The transcript id, not the session name — names are not stable. On 2026-08-23 one session was
`platform-e2` and then `platform-18`, and another `platform-59` then `platform-72`, both renamed by a
restart in a single night. The transcript id survived both.
Why it earns a line in every commit: three agents committed to this repository in one night and `git log`
could not tell them apart. `Co-Authored-By` says "an agent", which stopped being useful the moment there was
more than one. And the transcript holds the reasoning a commit message cannot — the arguments considered and
dropped, what was measured, what the owner decided and why. A commit that names its session is a commit
whose "why" is still answerable in six months.
**The precondition, and it is not built yet:** an agent cannot reliably discover its own transcript id. The
only exposure is that a background task's output path contains it, which means starting a job to learn who
you are. `commanderBlock` already hands a commander its id in the preamble, so the mechanism exists —
`todo/agent-coordination/` carries giving it to every session. Until then, an agent that knows its id should
add the trailer and one that does not should leave it out rather than guess.
## Frontend route conventions (apply to EVERY new dashboard route)
Worked examples: `/soulseek`, `/music`, `/chat`.
- **Workspace/Panel framework, always — never a standalone single screen.** The screen renders
`<WorkspaceView workspace={ws} locked />` where
`ws = useDashboardState<LayoutNode>('screens/<name>', defaultLayout)`, guarded by a `normalizeLayout`
that pins `appType`s to an allow-list. Panels are windowed apps under
`src/workspaces/officerdev/src/apps/<Feature>/`, each exporting `appRegistryMetas`
(`{ key, name, icon, component, availableOnPanel: false }`) and registered in `AppRegistry.tsx`.
The Workspace/Panel framework is orthogonal to routing — it contains no route navigation, and panels
live inside the Route element tree, so they can call `useParams`/`useSearchParams` directly.
- **Page title by route.** Automatic — no rule to add. `titleForPath` in
`src/apps/officer-web/state/usePageTitle.ts` takes the URL's first path segment, turns `kebab-case`
into `Title Case`, and that's the tab title; `usePageTitleSync` does the rest. A screen that wants to
say more than its route does (a chat's own name, not just "Chat") publishes an override instead —
`usePublishPageTitle` in `officerdev`.
### The URL is the source of truth for selection
`docs/navigation-audit.md` is the authority here — read it before building a screen that selects
things. It names the **"opaque click" anti-pattern**: an element that opens something addressable but
keeps the id in an onClick closure instead of the DOM, leaves the URL unchanged, holds the selection
in `usePanelChannel`/`useGlobal`, and has no anchor semantics (no cmd-click, no middle-click, not
link-focusable). Half the app still does this; none of the new code should.
- **Addressable state goes in the URL** (`useParams` / `useSearchParams`), never in a channel.
`usePanelChannel` is for genuine signals and refresh buses (`files:refresh-signal`,
`SLSKD_REFRESH_CHANNEL`, `MUSIC_RESYNC_CHANNEL`) — and declare one with `defineChannel` in
`officerdev/src/channels.ts` rather than spelling the name and type at each site.
"Which thing is open" is a URL. Panels each read the URL rather than passing it between themselves.
- **Rows and nav items are real links.** `<Link>` for rows (exemplar: the `/chat` session list,
`f35c145`); **react-router's `<NavLink>`** for nav chrome, so active state comes from the router.
The hand-rolled `isActive` in `Dock`/`Header` is scheduled for replacement (audit Phase 4) — don't
copy it. A disabled entry renders as a `<span>`; a disabled `<a>` is not a thing. A control that
_mutates_ rather than navigates stays a `<button>`.
- **Route pairs.** A bare screen route plus a param route rendering the same component: `/chat` +
`/chat/:sessionId`, `/jobs` + `/jobs/:id`, `/headscale` +
`/headscale/:section`. One `<Navigate … replace />` guard in the screen, placed after all hooks,
canonicalises both the bare route and a bogus param.
- **Master list + live preview uses `?selected=<id>`**, not the detail route — linking rows straight
to `/x/:id` destroys preview-on-list, because that route is the full page. Rows link to
`/x?selected=id` on-page and `/x/:id` off-page; list, mobile panel and preview all read the param.
Action buttons are **siblings** of the anchor, never nested inside it.
- **`src/workspaces/components/NavLink.tsx` is not react-router's `NavLink`** — it's a `<Link>`
wrapper that appends `useGlobalQueryString()` and provides no active state. Import react-router's
when you want `isActive`.
- Route helpers shared between an `officerdev` panel app and an `officer-web` screen must be
re-exported from `src/workspaces/officerdev/src/index.ts` (named exports only — the barrel
deliberately avoids `export *` for app modules to keep `appRegistryMetas` from colliding).
## COMMS — a channel between agents, when one is open
`COMMS/<work-stream>/` is a **tracked** channel between agents working on this repo from different machines.
It exists because findings used to reach each other by the owner relaying them from memory at the end of long
sessions.
**There is no open channel right now.** `COMMS/sidecar-app-store/` ran for one night — per-user Linux
accounts through to a member's first agent turn — and was deleted when the work landed, which is the
convention rather than an oversight: a spent channel left in place gets read as current.
If you open one:
- **Read it before starting**, if your task touches its work stream. It carries what is verified, what is
assumed, what is broken, and what is waiting on a decision — the parts a commit message does not hold.
- **Number the files and alternate**, one per turn, odd for one agent and even for the other. The parity is
the author; the alternation is the protocol. A push with no doc is then visibly a break rather than
something to find by diffing, and "nothing to report" is still a turn worth taking — silence and a crashed
agent read identically.
- **End on a checkable condition**, not on either party's judgement: no open item is actionable by a
participant. "I think we're done" can close a thread with work still in it.
- **Durable reasoning goes in `docs/` or next to the code.** The channel is for coordination. When the work
lands, delete the channel and move anything still open to `TODO.md`.
Two things that made it work, and neither is about either agent being more careful. One writes, the other
verifies, and only the verifier runs things on a real machine — most of what was caught was invisible to
reading and needed a live filesystem. And the author of a comment is the worst-placed person to notice the
code disagrees with it: the two most serious defects were both found by whoever had not written the sentence
explaining why it was safe.
## Further Reading
- `docs/navigation-audit.md` — **authoritative** on routing/navigation: the opaque-click anti-pattern,
a severity-ranked findings table, the channel-selection map and the four-phase plan
- `docs/email-recall.md` — **design, not built**: recall over your own correspondence. The split that
matters is counting (needs every message judged) versus retrieval (needs the best few, fast) — and why
a number produced by searching is confidently wrong with no signal that it is
- `docs/find-my-device.md` — **design, not built**: locating your own devices from Officer. What is
assembly (most of it), what is impossible (offline finding — there is no crowd to relay it), and the
question to settle first, which is retention rather than anything technical
- `docs/affordances.md` — what an affordance is, the three tests for whether something qualifies (including
the one that catches most bad ones: _can it be fixed instead?_), the file format, and how they reach the
agent. Read before writing one
- `docs/agent-coordination.md` — **the north star** for the workspace/panel work: agents on one
dashboard coordinating with each other instead of through the human, the handoff protocol, and what
is deliberately _not_ being built. Read it before ranking, deferring or starting any panel item —
it is what `docs/workspace-panel-todo.md` is ranked against.
- `docs/workspace-panels.md` — how the Workspace/Panel framework works: the layout tree, how a panel is
mounted and what it knows, the URL-vs-channel split for panel-to-panel communication, and the
persistence key families. Read before building a panel app. Its defect list is
`docs/workspace-panel-todo.md`.
- `docs/multi-server-push.md` — **design, not built**: one Officer app receiving notifications from every
server it is signed into. Works already, bar one missing field — and why the `mv` catalogue-version
check turns into an endless resync between servers without it. The server identifier is an SSO decision
- `docs/visual-regression-plan.md` — **plan, not built**: nightly screenshot regression against a JWT
dashboard on a second server. The blocker is that a JWT lives in `localStorage` and iris copies only
cookies, so an authenticated capture silently renders the login page — plus why diffing must happen
before any agent is asked to look
- `docs/plugin-dependency-scope.md` — a plugin declares WHO must be able to run each host binary. Why the
owner's login shell is the wrong question for a `confined` plugin, the two-part reachability check
(world-executable AND every parent traversable), and why members' binaries on alpha worked only because
one home directory happened to be mode 751
- `docs/plugin-route-authorization.md` — **design, not built**: making a plugin's settings routes
owner-only while the plugin itself stays grantable. What the proxy should pass (`isOwner` and a level,
never the role name), why declaring owner-only paths in the manifest beats trusting every plugin to
remember, and the credential-resolution work that has to land first
- `docs/roles-and-teams.md` — **plan, phase 0 built**: roles as a table, teams as scopes, and the one
change that has to land first — the owner becomes an identity rather than whoever holds a role name,
because otherwise "add a role" and "become the owner" are the same feature
- `docs/localization.md` — **plan, not built**, with a working 5% skeleton already in the tree: taking the
web UI to more than one language. What exists and what it cannot do yet, why the server should return
error CODES rather than prose (and why a code must not leak more than the prose it replaced), the key
naming convention, and the four phases. Read before adding a user-facing string
- `docs/secret-store.md` — **design, not built**: moving the encryption and signing keys out of `.env`
into a SQLite store, why they cannot live in Postgres, and key rotation. Also records the core/plugin
split it assumes — light plus `officer-headscale` is the core; Vaultwarden and the wallet are plugins
- `docs/sidecar-topology.md` — where the sidecar architecture is going, and what was considered and dropped
- `docs/working-on-officer.md` — how to run, restart and check your work on this machine
- `docs/wallet-key-custody.md` — what the platform can and cannot see of the wallet
- `docs/editor-buffer-corruption.md` — an open bug with a trap set for it: background-job output appearing
inside a file open in the editor. Read it BEFORE investigating a report of that, and before re-testing
anything it already rules out
- `docs/commander-directory.md` — **designed, not built**: a commander is a directory, not a session. Why
its identity is currently borrowed from Claude's transcript id and what that costs — an ordering problem,
a first-turn interlock that is a side effect rather than a signal, and a commander that dies with its
conversation. The CLAUDE.md in its own working directory is the living document, for free
- `docs/claude-md-touchpoints.md` — **audit**: every place the platform writes, appends to or reads a
`CLAUDE.md` — three writers total (commander creation, the plugin chat-brief copier, and a one-off
credential-backfill migration) plus one read-only doc-drift check, and three places that deliberately
do NOT write one. Read before changing any part of the mechanism; includes a per-plugin table (only
`proxy-manager` ships a chat brief today)
- `docs/plugin-capabilities.md` — **built**: a plugin's tasks, skills and affordances are
READ where they live, never copied into `capabilities/`. Why copying breaks a forkable repository, why it
goes stale on update and survives a disable, and why separate directories beat namespaced paths —
a namespace makes provenance readable, separate directories make it structural
- `docs/filesystem-quotas.md` — **planned, not done**: the host-side half of per-user and per-team storage
limits. Why the kernel has to enforce it rather than the platform (members have `terminal` and can write
past anything we count), why `usrquota`+`grpquota` and not project quotas, and the one thing easiest to
get wrong — a team folder that is not `setgid` charges nothing to the team
- `TODO.md` — current direction and deferred work; **takes precedence over this file where they disagree**
- `src/apps/CLAUDE.md` — shared frontend patterns
- `src/databases/CLAUDE.md` — database patterns
- `CONVENTIONS.md` — component organisation, state management and React patterns, with rationale;
`src/workspaces/officerdev/APP_CONVENTIONS.md` and `HOOK_CONVENTIONS.md` for panel apps and hooks
Treat the code as the source of truth where anything here disagrees with it.
+3 -560
View File
@@ -1,563 +1,6 @@
# CLAUDE.md # CLAUDE.md
## Project Overview **The instructions live in [`AGENTS.md`](AGENTS.md). Read it first — this file only points there.**
Officer is a self-hosted platform built around one person — the server owner — which since 2026-08-07 Kept separate so agents that look for `AGENTS.md` by convention find the same guidance as those that
also admits **additional accounts holding a strict subset of it**. It bundles an AI agent, a terminal, look for `CLAUDE.md`, without the two drifting apart.
a file browser, a code editor, a bitcoin wallet, a remote desktop and customisable dashboards
behind a single web app.
**The owner/member split, and where the line falls.** This file said "single-user is a hard invariant,
not a stage" until 2026-08-07. That is no longer true and had already stopped being true when it was
written: `users` holds six rows. The accurate statement is narrower and more useful —
- **One owner.** User id 1, role `Super Admin`, created by `POST /auth/bootstrap` while the table is
empty, pinned there by a CHECK constraint. The owner bypasses every permission check.
- **Other accounts get only what their ROLE is granted.** Roles are `Admin`, `Member`, `Developer`;
grants live in `role_permissions`, keyed on role, never on user. Absence denies — there is no row
meaning "no", so an empty table is a server where members reach nothing but their own profile.
- **Some things can never be shared, structurally.** Tasks, items, desktop and browser are
`kind: 'execution'`: they run as the owner's OS user in the owner's home, so there is no level of
"read" that makes them safe. They have no level at all and the grants API refuses to store one.
- **And some are shared only because the kernel enforces it.** Terminal, chat and files are
`kind: 'confined'`, added 2026-08-11 with per-user Linux accounts. They still touch the filesystem
and still run processes — but not the _owner's_, because the account has its own Linux user, its own
home, and the kernel refusing everything above it.
The distinction earns its keep in one place: **a confined grant means nothing without that Linux
user.** `authorize.ts` drops it for an account whose `osUser` is null, so "granted but unconfined"
resolves to no access rather than to the owner's home — which is what it would otherwise resolve to,
since `getOwnerHomeDir` ignores the email it is passed. That rule lives there once and covers the
HTTP routes, the websocket doors and the dock together.
So "which user is this" has a real answer for the **app** surface (gitea, music, photos,
calendar…) and for the **confined** one (terminal, chat, files), and is still always "the owner" for
anything under `execution`.
`src/servers/permissions/registry.ts` is the authority and reads as the design document for this.
**Mounting a router without a registry entry makes the server refuse to boot** — see "Permissions"
below before adding one.
**Account creation exists.** `POST /api/users` (owner-gated), with a form under Settings → User Management:
it validates the password with the same rules bootstrap uses, refuses a role that is not a row in `roles`,
refuses `Super Admin` outright because there is exactly one owner, optionally takes an inbound SSH key, and
provisions the Linux account. `POST /users/:id/provision-linux` retries that half if the host was not ready.
This file said "there is no way to create a second account" for some days after `69a31051` shipped one, and
`docs/roles-and-teams.md` sequenced a whole phase around building it. Neither is checkable by
`bun run docs:check` — it compares numbers, and this was prose. Worth knowing as the shape of the failure:
the counts stay right and the sentences rot.
## Architecture
One Bun process (`src/server.tsx`) serves everything:
- the React SPA, via Bun's HTML import of `src/apps/officer-web/index.html` (HMR in dev)
- the REST API, a Hono app mounted at `/api` (`src/servers/hono.ts`)
- eight WebSocket providers — terminal, chat, task-runner, pipeline, cliamp, cliamp-audio, desktop,
vault — plus a sidecar registration socket. `terminal` is a byte relay onto the pty sidecar's own
listener, not a translating bridge; `vault` is the same shape onto Vaultwarden's notifications hub.
- ~~a browser relay on its own port~~ — switched off 2026-08-13, awaiting extraction into a plugin.
The extension and `api/browser/` stay on disk; the listener and the `/api/browser` mount do not.
Long-running and privileged work lives in **sidecars**: separate processes that dial back in over
`/api/sidecar/register` and are tracked in `src/servers/sidecar-registry.ts`. PM2 runs them
(the generated `ecosystem.config.cjs` — see below): `officer` (the server), `officer-anthropic-proxy`, `officer-claude-code`,
`officer-opencode`, `officer-pty`, `officer-vnc`, `officer-music`, `officer-vault`,
`officer-slskd`, `officer-headscale`, `officer-transmission`, `officer-invoiceshelf`, `officer-wallet`,
`officer-photos`, `officer-notify`, `officer-caldav`, `officer-memos`, `officer-jellyfin`, `officer-gitea`
— twenty as of 2026-08-06, and a list that goes stale every time a sidecar lands. `pm2 jlist` is the
source of truth.
**`officer-anthropic-proxy` and `officer-claude-code` are not the same thing.** (The second was
called `officer-agent` until 2026-08-13; older docs use that name.) The proxy holds the Anthropic
credential and forwards API traffic; the agent is the process that spawns `claude`. They were one entry
named `officer-claude` until the sidecar-isolation work — which is exactly how the false claim that
"restarting officer doesn't disturb the agent" survived so long. Every sidecar is a PM2 peer of
`officer`, so **no sidecar is a child of the server and restarting the server does not kill one.**
Agents run **unsandboxed as the server owner**, with `--dangerously-skip-permissions`. This is
deliberate — it is the owner's own machine. Do not add a jail without being asked.
## Repository Layout
```
src/
├── server.tsx # the single entrypoint: routes, WS upgrades, static files
├── apps/
│ └── officer-web/ # the SPA shell — Screens/Authentication + Screens/Dashboard
├── servers/
│ ├── hono.ts # router composition; everything under /api
│ ├── _middlewares/ # auth, body parsing, the permission gate, rate limiting
│ ├── api/<feature>/ # one folder per feature, each exporting a router
│ ├── channels/ # send-claude-code / send-opencode — how /chat drives an agent turn
│ ├── queue/ # background job engine
│ └── sidecar/ # sidecar implementations + the wire protocol
├── databases/officer_db/ # the only database (Postgres + Drizzle)
├── extensions/browser-relay/
└── workspaces/ # shared packages, each a bun workspace
```
`src/workspaces/officerdev` is the biggest of these: the windowed "apps" (FileBrowser, Chat,
Terminal, CodeEditor, Desktop, Dashboards, Wallet…) that the shell hosts, behind an `AppRegistry`.
`src/apps/officer-web` is only the shell — screens, routing and settings.
**Path aliases** (`tsconfig.json`): `@/*` → `src/apps/officer-web`, `@/components/*` →
`src/workspaces/components`, `@@/*` → `src/servers`, `@/public/*` → `public`. Workspace packages are
imported by their package name (`officerdev`, `hooks`, `state`, `types`, `helpers`, …).
## Tech Stack
- **Runtime**: Bun (Node 22 or newer is enforced by a `preinstall` check)
That check demanded _exactly_ 22 until 2026-08-12. The reason was a `node-pty` build
failure some months earlier, whose details were not recorded. It was relaxed to `>= 22`
after confirming node-pty ships **no Linux prebuilds** — its install script always falls
through to `node-gyp rebuild`, so it compiles against whatever Node is present and there
is no ABI to mismatch. Untested on 24 at the time of the change. If `bun install` fails
building node-pty, or `officer-pty` cannot load its native module, restore the exact pin
first. The source build also needs `build-essential` and `python3`.
**Bun itself is held at the 1.3 series, on purpose and temporarily.** 1.4.0 was published
2026-08-20; the owner asked on 2026-08-22 to stay on 1.3 for a few months until 1.4 has been
watched. Two places carry that number and they move together: `BUN_SERIES` in
`scripts/setup/machine-setup/lib/dev.sh`, which resolves the newest `1.3.x` from the npm registry
and passes it to the installer as a release tag, and `engines.bun` in `package.json`, which is
what `bun run doctor` reads. bun ignores `engines` (verified — an install declaring `1.9.x` ran
clean under 1.3.14), so that field is the declaration and doctor is the enforcement.
**Do not run `bun upgrade` on a machine running Officer.** It takes no version argument — latest
stable or `--canary`, nothing else — so it jumps straight past the pin and says nothing. That is
the whole reason `doctor` checks the running version at all; `machine-setup.sh --only 'JavaScript
runtimes'` offers to pin a drifted machine back down. Lifting the hold is `todo/infra/`.
- **Language**: TypeScript, strict. `bunx tsgo` is clean — keep it that way.
- **Frontend**: React 19, React Router 7, React Query, Tailwind 4, shadcn/ui + custom components
- **Backend**: Hono
- **Database**: Postgres via Drizzle — `officer` for the platform, plus `officer_marketplace` on the
host that serves marketplace.officer.dev (see `docs/marketplace.md`)
- **Agents**: Claude Code and opencode, driven through sidecars; tools exposed over MCP
## Data
Two stores, and the split matters:
**Postgres** (`src/databases/officer_db`) holds the account, passkeys, settings, dashboards,
queue and pipeline jobs. Since 2026-08-16 there is a SECOND database,
`officer_marketplace` (`src/databases/marketplace_db`), and the split is deliberate: it is a public
catalogue served by a process with no login, no session and no user, so sharing a connection would put
an unauthenticated surface one query away from every private table. It runs only on the host serving
marketplace.officer.dev — `docs/marketplace.md`. One directory per feature holding `schema.ts` and
`queries.ts` beside each other; `src/schema.ts` is what `db:push` reads, and it lists the core tables
with the plugin ones commented out. Types inferred from the schema in `src/types.ts`.
**The filesystem** holds everything the agent authors. `OFFICER_ITEMS_DIR` (`$OFFICER_ROOT/capabilities`)
contains one directory per item under `skills/`, `tools/`, `tasks/`, `processes/`, `extensions/` — no
database rows, no scope tiers. `DATA_PATH/<email>/` holds the managed home and attachments. It also holds
`email_accounts/`, which is the **email plugin's** — the platform pre-creates the directory and never opens
what is inside it.
**None of those paths is configured.** Since 2026-08-13 `src/servers/data-path.ts` derives the install
root as `resolve(process.cwd(), '..')` and hangs `data/`, `permissions/` and `dockers/` off it. That
replaced `DATA_PATH`, `OFFICER_ITEMS_DIR` and `HOME_DIR` in `.env` — three values that had to agree with
each other and with the tree on disk. `assertInstallLayout` refuses to boot when the working directory
is not the repo, because otherwise a wrong `cwd` relocates the whole install silently rather than
failing.
Note `getHomeDir` (the managed home under `DATA_PATH`, now used only for NON-owner accounts and by
pipeline-executor) versus `getOwnerHomeDir` (the owner's real login home, where terminals, chats and
task runs execute — captured from `homedir()` once at module load, and it ignores the email it is
passed).
### Schema changes use `push`, not migrations
This database is kept in sync with `bun db:push`, which diffs the schema code against the live
database and alters it directly. **`drizzle-kit migrate` has never been run here** — there is no
`__drizzle_migrations` table. Change the schema, run `bun db:push`, done.
`bun db:gen` writes files to `officer_db/migrations/`, but nothing applies them; the old numbered
history was deleted because it had drifted from the real schema. Treat the schema code, not those
files, as the source of truth.
**Declare multi-column uniqueness as `uniqueIndex('uq_…').on(a, b)`, never `unique('uq_…').on(a, b)`** —
drizzle-kit mis-diffs named composite unique _constraints_ and re-creates them on every push, which used
to stop `db:push` on an unanswerable truncate prompt. Same for any foreign key whose generated name would
exceed Postgres's 63-character identifier limit: name it explicitly. See `src/databases/CLAUDE.md` →
"Composite keys" before adding either.
## Security Model
- `IS_DEV_BUILD` (`src/servers/build-env.ts`) is true **only** when `PUBLIC_BUILD_ENV` is explicitly
`dev`/`development`. Everything else, including unset, is hardened. Rate limiting and password
rules key off it — they fail closed.
- **There is no origin checking.** It was removed on 2026-08-13, along with `ALLOW_ANY_ORIGIN` and
`ALLOW_ANY_ORIGIN_MUSIC`. The flag defaulted to ON, so origin validation ran on no real install —
what came out was documented defence in depth that was already switched off. Origin was never
authentication here anyway: an app's `officer://<hex>` origin is chosen by the client, forgeable
outside a browser, and extractable from a shipped binary. The perimeter is the tailnet, and the lock
is a valid token on every protected route plus the permission gate below.
- JWTs are 30-day, blacklisted on signout, and invalidated by a password change (`passwordChangedAt`).
**The role is deliberately not a claim** — every authorization decision re-reads `users.role` from
Postgres, so a grant or a revoke takes effect on the next request rather than at next sign-in.
- A panic lockdown (`src/servers/api/auth/panic.ts`) is in-memory only and refuses every
authenticated request until the server restarts.
### Permissions — read this before mounting a router
Authorization is one system, and it is not in `userMiddleware` (which only answers "is this token
valid"). It is `_middlewares/permission-gate.ts` → `permissions/authorize.ts`, mounted globally in `hono.ts`
ahead of everything, and it re-verifies the token itself so it covers routes that never mount
`userMiddleware`.
- `permissions/registry.ts` — the single enumeration of what the platform can do, in five kinds:
`core` (every account, not deniable), `app` (**the grantable surface**), `confined` (grantable, but
only to an account that has a Linux user), `execution` and `admin` (owner only, and `execution` is
never grantable at any level). **21 entries as of 2026-08-27** — verified by `bun run docs:check`, which fails if this number drifts again: 6 core, 2 app, 4 confined, 3 execution, 6 admin (`todos`, the todo board, is the sixth). It was 22 until caldav became a plugin and took `calendar` with it — a plugin's permission is registered at runtime from its manifest, so it is real but not in this file's count. `notify` moved from `app` to `core` on 2026-08-24: it was grantable, nothing granted it, and every Member's app was silently 403'd registering its own push token. It said 27 for nine days. The execution count dropped by one when Minesweeper became a plugin — see `plugins/minesweeper/manifest.ts` for why that was a fix rather than a downgrade.
- `permissions/authorize.ts` — resolves "may this account do this". Owner short-circuits first; every
other answer is role grants plus core, with `execution`/`admin` stripped even if a row grants them,
and `confined` stripped for an account with no `osUser`.
**Every catch returns deny.** Grants are cached by role and the cache's whole invalidation contract
is `invalidateRoleGrants`, called by the one writer in `api/users/permissions-routes.ts`.
- `permissions/totality.ts` — `assertPermissionTotality` runs in `server.tsx` **before `serve()` and
throws**. Mount a router or a socket without a registry entry and `pm2 restart officer` fails,
naming what is missing. That is deliberate: the hole it closes was a Member 403'ing on
`GET /api/tasks` and opening `/api/tasks/pipeline/ws` with a 101 in the same minute, because Bun's
route table matches the socket before the `/api/*` catch-all that reaches Hono. A patch does not
survive the next door; refusing to boot does.
So **adding a router means adding one line to `CAPABILITIES`**. If the surface genuinely is not
user-gated, add it to `EXEMPT_API_PREFIXES` in `totality.ts` _with a reason_ — an unexplained exemption
is how the hole happened the first time.
The frontend hook `usePermissions` **fails open** on purpose: hiding a dock icon is a courtesy, the
403 is the lock, and an owner locked out by a transient network error is worse than a member clicking
into a refusal.
## Commands
```bash
bun dev # the whole app — SPA + API + WebSockets, watched
bun start # production
bunx tsgo # typecheck (not tsc)
bun test # tests
bun format # prettier over every dirty file — see the note below before running it
bun db:push # apply the schema to Postgres
bun docs:check # do the docs still match the code? REQUIRED before tagging a release
bun setup # runs scripts/install.sh — blank machine to running platform
```
**`bun docs:check` before every release, and leave it green.** It compares documented claims against values
derived at run time — the permission count, the PM2 entry count, every settings route Spotlight offers — and
exits non-zero on drift. A check must never compare a doc against a second hard-coded copy; that is one more
place to drift, and it drifts in the same edit that misses the doc.
It is narrow on purpose. It cannot read prose and judge it, only claims with a machine-checkable
counterpart — which is a small part of the documentation and the part that rots fastest, because a number is
exactly what nobody re-derives while editing the sentence around it. Both claims it found on the day it was
written had been wrong for over a week.
`scripts/install.sh` is only an orchestrator — it runs the two halves in order and does nothing itself:
`setup/machine-setup/machine-setup.sh` (28 sections: packages, tailnet, runtimes, docker, shell) then
`setup/officer-setup.sh` (11: pre-flight, layout, repository, dependencies, database, environment, secrets,
schema, build, services, verify). Either runs alone — `--machine-only`, `--officer-only`, or by path — because
a machine you already trust needs only the second. Both are re-runnable: each records the steps it finished
and skips them, so stopping halfway costs nothing. **Run it as yourself**; it re-execs through `sudo` when it
needs to, and on macOS never does, because Homebrew refuses to run as root.
Sidecar control is PM2, not npm scripts: `pm2 restart officer-<name>`, `pm2 logs officer-<name>`.
See `docs/working-on-officer.md` for which process a given change needs restarted.
### Installs are frozen. Never resolve a dependency you did not ask for.
`bunfig.toml` sets `[install] frozenLockfile = true`, so **`bun install` resolves from `bun.lock` and
nothing else** — it fails rather than quietly picking up a newer version, including a transitive one
nobody chose. Verified on bun 1.3.10: a lockfile that no longer satisfies `package.json` exits 1 with
`error: lockfile had changes, but lockfile is frozen`.
It is config rather than a habit because a supply-chain compromise does not wait for the one time
somebody forgets a flag. On **2026-08-04** eleven cache packages — `keyv`, `flat-cache`,
`file-entry-cache`, `cacheable-request`, `cache-manager`, the `@cacheable/*` scope, `ecto` — were
published with a `preinstall` dropper that harvested npm and GitHub tokens, AWS and Kubernetes
credentials, SSH and PEM keys, `.env` files and `.claude/settings.json`, then republished itself
through any npm token it found. It reached 434 further packages across 1,381 versions. This machine was
unaffected only because nothing had installed since 2026-08-02.
**To change a dependency:** edit `package.json`, allow the lockfile to change deliberately, **read the
diff**, and commit it. Note that `--no-frozen-lockfile` — which this file and `bunfig.toml` both used to
name — is NOT a flag bun 1.3.14 has; it is silently ignored and the install still fails frozen. Comment
out `frozenLockfile` in `bunfig.toml` for the one command, then put it back. The friction is the point — an unexplained lockfile change
in a diff is the signal this exists to produce.
**Do not** add `--no-frozen-lockfile` to a script, a Dockerfile or CI to make an error go away. The
error means the lockfile and `package.json` disagree, and that is worth a human look every time.
**Format your own files, not the whole dirty tree.** `bun format` globs `git diff --name-only HEAD`,
so it rewrites every uncommitted file — including work in progress that isn't yours, which then shows
up as unexplained whitespace churn in someone else's diff. Run `bunx prettier --write <paths>` on the
files you actually touched. `bun format` is only safe when the tree is otherwise clean.
## Code Style
- **Paradigm**: functional — pure functions, immutability, composition
- **TypeScript**: strict, no `any`. Type-only imports are required (`verbatimModuleSyntax`).
- **Comments**: minimal, and about _why_. Don't narrate what the code already says.
- **Async**: always async/await
- **Exports**: named only, no defaults
- **Files**: `PascalCase.tsx` for components, `kebab-case.ts` for everything else
- **Prettier**: single quotes in JS, double in JSX, semicolons, trailing commas, 120 cols
### Imports
Order: types → external → workspace → relative.
```ts
import type { User } from 'types';
import { useState } from 'react';
import { useQuery } from '@tanstack/react-query';
import { useClient } from 'hooks/useClient';
import { formatDate } from 'helpers/formatters';
import { useWebsites } from '../useWebsites';
```
In app code import types from `'types'`, which re-exports the database types. Only server and
database code imports from `officerdb/types` directly.
### Functions
Arrow functions for one-liners, regular functions for anything with real logic. No multiline
parameter lists — extract a params type instead.
```ts
type ProcessDataParams = { body: Record<string, unknown> | undefined; query: Record<string, string> };
export function processData({ body, query }: ProcessDataParams): Result { ... }
```
### Components and hooks
```tsx
type ExperimentCardProps = { experiment: Experiment; onSelect: (id: number) => void };
export const ExperimentCard = ({ experiment, onSelect }: ExperimentCardProps) => { ... };
```
Complex hooks return objects; simple state hooks return `as const` tuples. Event parameters are
always named `ev`, never `e`.
### API calls
`useClient()` returns typed verbs — `client.get<Foo>('/foo')` really does return `Foo`, so pass the
type parameter and let inference flow from it.
## Working With Me
- **Ask first** — confirm the approach before a significant change
- **Explore thoroughly** — read the related files before editing
- **Keep it simple** — no over-engineering, no premature abstraction
- **Be explicit** — no magic, no implicit behaviour
- **Stay focused** — note unrelated problems, don't fix them uninvited
- **Report honestly** — say what you verified and what you didn't
**Always commit and push when you finish implementing — including when it went wrong.** Don't wait to be
asked, and don't hold a branch back because it is unfinished, untested or turned out to be a dead end. The
history of the mistakes is worth having: a reverted commit and its message explain why an approach was
abandoned, which is exactly the thing that gets lost when a failed attempt is quietly discarded. Say what
state it is in — in the commit message and in `COMMS/` if another agent will pick it up — rather than
withholding the commit until it is good.
Commit messages: simple lowercase, no prefixes.
**Stage and commit in ONE step, naming the paths.** `git add <path>` then a bare `git commit` commits the
whole **index**, not the path just named — so anything a concurrent session had staged goes into your
commit, under your message and your session trailer. Use `git commit -- <paths>` instead, which does what
the two-step form only looks like it does.
This is not hypothetical and it is not rare: two agents share this working tree most nights. On
2026-08-23 commit `6e092c32` says "marketplace: minesweeper in the catalogue" and contains nine files of
another session's roles work, swept up mid-`git add`. The other session was right not to amend it —
rewriting shared history to fix a message is the worse trade — so the record was corrected with a
follow-up note instead. Cheap to avoid, awkward to undo.
Two things about the form itself, both learned the same night. `-m`/`-F` goes BEFORE the `--`, or git
reads the message as a pathspec. And it only works on **tracked** files — a new file must be `git add`ed
first, so a commit introducing one is `git add <new>` then `git commit -- <paths>`, naming every path.
Getting that wrong fails loudly, which is the good case: the release tag that followed it went onto the
wrong commit, and only because the tag was checked afterwards did that surface at all.
**An agent's commit names the session that made it.** A trailer, below `Co-Authored-By`:
```
Officer-Session: 656f5c85-e49a-4f1f-b9f3-042a04b4d043
```
The transcript id, not the session name — names are not stable. On 2026-08-23 one session was
`platform-e2` and then `platform-18`, and another `platform-59` then `platform-72`, both renamed by a
restart in a single night. The transcript id survived both.
Why it earns a line in every commit: three agents committed to this repository in one night and `git log`
could not tell them apart. `Co-Authored-By` says "an agent", which stopped being useful the moment there was
more than one. And the transcript holds the reasoning a commit message cannot — the arguments considered and
dropped, what was measured, what the owner decided and why. A commit that names its session is a commit
whose "why" is still answerable in six months.
**The precondition, and it is not built yet:** an agent cannot reliably discover its own transcript id. The
only exposure is that a background task's output path contains it, which means starting a job to learn who
you are. `commanderBlock` already hands a commander its id in the preamble, so the mechanism exists —
`todo/agent-coordination/` carries giving it to every session. Until then, an agent that knows its id should
add the trailer and one that does not should leave it out rather than guess.
## Frontend route conventions (apply to EVERY new dashboard route)
Worked examples: `/soulseek`, `/music`, `/chat`.
- **Workspace/Panel framework, always — never a standalone single screen.** The screen renders
`<WorkspaceView workspace={ws} locked />` where
`ws = useDashboardState<LayoutNode>('screens/<name>', defaultLayout)`, guarded by a `normalizeLayout`
that pins `appType`s to an allow-list. Panels are windowed apps under
`src/workspaces/officerdev/src/apps/<Feature>/`, each exporting `appRegistryMetas`
(`{ key, name, icon, component, availableOnPanel: false }`) and registered in `AppRegistry.tsx`.
The Workspace/Panel framework is orthogonal to routing — it contains no route navigation, and panels
live inside the Route element tree, so they can call `useParams`/`useSearchParams` directly.
- **Page title by route.** Automatic — no rule to add. `titleForPath` in
`src/apps/officer-web/state/usePageTitle.ts` takes the URL's first path segment, turns `kebab-case`
into `Title Case`, and that's the tab title; `usePageTitleSync` does the rest. A screen that wants to
say more than its route does (a chat's own name, not just "Chat") publishes an override instead —
`usePublishPageTitle` in `officerdev`.
### The URL is the source of truth for selection
`docs/navigation-audit.md` is the authority here — read it before building a screen that selects
things. It names the **"opaque click" anti-pattern**: an element that opens something addressable but
keeps the id in an onClick closure instead of the DOM, leaves the URL unchanged, holds the selection
in `usePanelChannel`/`useGlobal`, and has no anchor semantics (no cmd-click, no middle-click, not
link-focusable). Half the app still does this; none of the new code should.
- **Addressable state goes in the URL** (`useParams` / `useSearchParams`), never in a channel.
`usePanelChannel` is for genuine signals and refresh buses (`files:refresh-signal`,
`SLSKD_REFRESH_CHANNEL`, `MUSIC_RESYNC_CHANNEL`) — and declare one with `defineChannel` in
`officerdev/src/channels.ts` rather than spelling the name and type at each site.
"Which thing is open" is a URL. Panels each read the URL rather than passing it between themselves.
- **Rows and nav items are real links.** `<Link>` for rows (exemplar: the `/chat` session list,
`f35c145`); **react-router's `<NavLink>`** for nav chrome, so active state comes from the router.
The hand-rolled `isActive` in `Dock`/`Header` is scheduled for replacement (audit Phase 4) — don't
copy it. A disabled entry renders as a `<span>`; a disabled `<a>` is not a thing. A control that
_mutates_ rather than navigates stays a `<button>`.
- **Route pairs.** A bare screen route plus a param route rendering the same component: `/chat` +
`/chat/:sessionId`, `/jobs` + `/jobs/:id`, `/headscale` +
`/headscale/:section`. One `<Navigate … replace />` guard in the screen, placed after all hooks,
canonicalises both the bare route and a bogus param.
- **Master list + live preview uses `?selected=<id>`**, not the detail route — linking rows straight
to `/x/:id` destroys preview-on-list, because that route is the full page. Rows link to
`/x?selected=id` on-page and `/x/:id` off-page; list, mobile panel and preview all read the param.
Action buttons are **siblings** of the anchor, never nested inside it.
- **`src/workspaces/components/NavLink.tsx` is not react-router's `NavLink`** — it's a `<Link>`
wrapper that appends `useGlobalQueryString()` and provides no active state. Import react-router's
when you want `isActive`.
- Route helpers shared between an `officerdev` panel app and an `officer-web` screen must be
re-exported from `src/workspaces/officerdev/src/index.ts` (named exports only — the barrel
deliberately avoids `export *` for app modules to keep `appRegistryMetas` from colliding).
## COMMS — a channel between agents, when one is open
`COMMS/<work-stream>/` is a **tracked** channel between agents working on this repo from different machines.
It exists because findings used to reach each other by the owner relaying them from memory at the end of long
sessions.
**There is no open channel right now.** `COMMS/sidecar-app-store/` ran for one night — per-user Linux
accounts through to a member's first agent turn — and was deleted when the work landed, which is the
convention rather than an oversight: a spent channel left in place gets read as current.
If you open one:
- **Read it before starting**, if your task touches its work stream. It carries what is verified, what is
assumed, what is broken, and what is waiting on a decision — the parts a commit message does not hold.
- **Number the files and alternate**, one per turn, odd for one agent and even for the other. The parity is
the author; the alternation is the protocol. A push with no doc is then visibly a break rather than
something to find by diffing, and "nothing to report" is still a turn worth taking — silence and a crashed
agent read identically.
- **End on a checkable condition**, not on either party's judgement: no open item is actionable by a
participant. "I think we're done" can close a thread with work still in it.
- **Durable reasoning goes in `docs/` or next to the code.** The channel is for coordination. When the work
lands, delete the channel and move anything still open to `TODO.md`.
Two things that made it work, and neither is about either agent being more careful. One writes, the other
verifies, and only the verifier runs things on a real machine — most of what was caught was invisible to
reading and needed a live filesystem. And the author of a comment is the worst-placed person to notice the
code disagrees with it: the two most serious defects were both found by whoever had not written the sentence
explaining why it was safe.
## Further Reading
- `docs/navigation-audit.md` — **authoritative** on routing/navigation: the opaque-click anti-pattern,
a severity-ranked findings table, the channel-selection map and the four-phase plan
- `docs/email-recall.md` — **design, not built**: recall over your own correspondence. The split that
matters is counting (needs every message judged) versus retrieval (needs the best few, fast) — and why
a number produced by searching is confidently wrong with no signal that it is
- `docs/find-my-device.md` — **design, not built**: locating your own devices from Officer. What is
assembly (most of it), what is impossible (offline finding — there is no crowd to relay it), and the
question to settle first, which is retention rather than anything technical
- `docs/affordances.md` — what an affordance is, the three tests for whether something qualifies (including
the one that catches most bad ones: _can it be fixed instead?_), the file format, and how they reach the
agent. Read before writing one
- `docs/agent-coordination.md` — **the north star** for the workspace/panel work: agents on one
dashboard coordinating with each other instead of through the human, the handoff protocol, and what
is deliberately _not_ being built. Read it before ranking, deferring or starting any panel item —
it is what `docs/workspace-panel-todo.md` is ranked against.
- `docs/workspace-panels.md` — how the Workspace/Panel framework works: the layout tree, how a panel is
mounted and what it knows, the URL-vs-channel split for panel-to-panel communication, and the
persistence key families. Read before building a panel app. Its defect list is
`docs/workspace-panel-todo.md`.
- `docs/multi-server-push.md` — **design, not built**: one Officer app receiving notifications from every
server it is signed into. Works already, bar one missing field — and why the `mv` catalogue-version
check turns into an endless resync between servers without it. The server identifier is an SSO decision
- `docs/visual-regression-plan.md` — **plan, not built**: nightly screenshot regression against a JWT
dashboard on a second server. The blocker is that a JWT lives in `localStorage` and iris copies only
cookies, so an authenticated capture silently renders the login page — plus why diffing must happen
before any agent is asked to look
- `docs/plugin-dependency-scope.md` — a plugin declares WHO must be able to run each host binary. Why the
owner's login shell is the wrong question for a `confined` plugin, the two-part reachability check
(world-executable AND every parent traversable), and why members' binaries on alpha worked only because
one home directory happened to be mode 751
- `docs/plugin-route-authorization.md` — **design, not built**: making a plugin's settings routes
owner-only while the plugin itself stays grantable. What the proxy should pass (`isOwner` and a level,
never the role name), why declaring owner-only paths in the manifest beats trusting every plugin to
remember, and the credential-resolution work that has to land first
- `docs/roles-and-teams.md` — **plan, phase 0 built**: roles as a table, teams as scopes, and the one
change that has to land first — the owner becomes an identity rather than whoever holds a role name,
because otherwise "add a role" and "become the owner" are the same feature
- `docs/localization.md` — **plan, not built**, with a working 5% skeleton already in the tree: taking the
web UI to more than one language. What exists and what it cannot do yet, why the server should return
error CODES rather than prose (and why a code must not leak more than the prose it replaced), the key
naming convention, and the four phases. Read before adding a user-facing string
- `docs/secret-store.md` — **design, not built**: moving the encryption and signing keys out of `.env`
into a SQLite store, why they cannot live in Postgres, and key rotation. Also records the core/plugin
split it assumes — light plus `officer-headscale` is the core; Vaultwarden and the wallet are plugins
- `docs/sidecar-topology.md` — where the sidecar architecture is going, and what was considered and dropped
- `docs/working-on-officer.md` — how to run, restart and check your work on this machine
- `docs/wallet-key-custody.md` — what the platform can and cannot see of the wallet
- `docs/editor-buffer-corruption.md` — an open bug with a trap set for it: background-job output appearing
inside a file open in the editor. Read it BEFORE investigating a report of that, and before re-testing
anything it already rules out
- `docs/commander-directory.md` — **designed, not built**: a commander is a directory, not a session. Why
its identity is currently borrowed from Claude's transcript id and what that costs — an ordering problem,
a first-turn interlock that is a side effect rather than a signal, and a commander that dies with its
conversation. The CLAUDE.md in its own working directory is the living document, for free
- `docs/claude-md-touchpoints.md` — **audit**: every place the platform writes, appends to or reads a
`CLAUDE.md` — three writers total (commander creation, the plugin chat-brief copier, and a one-off
credential-backfill migration) plus one read-only doc-drift check, and three places that deliberately
do NOT write one. Read before changing any part of the mechanism; includes a per-plugin table (only
`proxy-manager` ships a chat brief today)
- `docs/plugin-capabilities.md` — **built**: a plugin's tasks, skills and affordances are
READ where they live, never copied into `capabilities/`. Why copying breaks a forkable repository, why it
goes stale on update and survives a disable, and why separate directories beat namespaced paths —
a namespace makes provenance readable, separate directories make it structural
- `docs/filesystem-quotas.md` — **planned, not done**: the host-side half of per-user and per-team storage
limits. Why the kernel has to enforce it rather than the platform (members have `terminal` and can write
past anything we count), why `usrquota`+`grpquota` and not project quotas, and the one thing easiest to
get wrong — a team folder that is not `setgid` charges nothing to the team
- `TODO.md` — current direction and deferred work; **takes precedence over this file where they disagree**
- `src/apps/CLAUDE.md` — shared frontend patterns
- `src/databases/CLAUDE.md` — database patterns
- `CONVENTIONS.md` — component organisation, state management and React patterns, with rationale;
`src/workspaces/officerdev/APP_CONVENTIONS.md` and `HOOK_CONVENTIONS.md` for panel apps and hooks
Treat the code as the source of truth where anything here disagrees with it.
+92
View File
@@ -0,0 +1,92 @@
# Plugins
`EXTRACTING-A-PLUGIN.md` is the runbook for turning a platform feature into a plugin, and
`EXTRACT-A-PLUGIN-PROMPT.md` is the way to hand that to a fresh agent. This file is for the concepts a
plugin author needs that are not procedure.
**This directory holds documentation, not plugins.** Every plugin is its own repository; `plugins/<name>/`
appears by `git clone` when somebody installs one, and is gitignored here.
---
## Affordances
**`docs/affordances.md` is the authority** — what one is, the three tests for whether something qualifies,
the file format and how they reach the agent. Read it before writing one. What follows is only the part
that is specific to shipping one from a plugin.
> An affordance is something the surface on the other end of the conversation offers the agent — declared
> by the code that implements it, and read by the agent at runtime. Not what the agent can do; what the
> client will honour.
A plugin ships an `affordances/` directory of markdown-with-frontmatter in its repository. Installing
copies it to `capabilities/affordances/<publisher>/<appName>/`, uninstalling removes it — the same shape
as `assets/` and `publishAssets`, and the namespace comes from where it lands rather than from anything
inside the file.
**The test that catches most bad ones: can it be fixed instead?** If the answer is yes, fix it. An
affordance describing a defect gives that defect a rationale, and the rationale outlives everyone's memory
that it was a bug. See the two worked examples in `docs/affordances.md`.
Keep it short — every affordance is loaded into every conversation on this machine, including turns that
have nothing to do with your plugin.
### A plugin-contributed one, for reference## Notifications — the same shape, for push
A plugin can also declare what its push notifications MEAN. Same mechanism, different store:
```
capabilities/notifications/
officerdev/music/
catalogue.yaml
```
```yaml
nowPlaying:
title: 'Now playing'
route: 'officer://music/{id}'
```
That becomes the code `officerdev/music:nowPlaying`. A push carries the code and integers — never content —
and the device renders the title and follows the route. The point is that **adding a notification is a
server change and never an app build**, because the mapping is synced to the device rather than compiled
into it.
**The namespace comes from where the file lands, not from what it says.** Writing `officer/core:test` as a
key in your own catalogue produces `<you>/<yourplugin>:officer/core:test`, not the core code — the same
rule as `mountPrefix`, and pinned by test in `src/servers/notifications/catalogue.test.ts`.
An entry needs a `title`; `body`, `target`, `route` and `thread` are optional, `thread` defaults to your
namespace, and unknown keys are dropped rather than forwarded to the device. Broken YAML in one plugin's
catalogue is skipped without affecting anyone else's.
See `docs/push-notifications.md`.
### A plugin-contributed one, for reference
This is the gitea plugin's, and it shows what these are for: the agent cannot discover from the
filesystem that `tea` is authenticated, or that it should prefer `git` for the things git already does.
```markdown
---
name: Gitea
description: Use when working with git repositories, or with Gitea issues and pull requests.
version: 1
---
For anything git already does — clone, fetch, pull, push, branches, tags — use plain `git`. It talks
to the server directly.
For anything git does not — pull requests, issues, comments, releases, creating a repository — use
`tea`. It is already logged in as the current user, with that user's own permissions. Do not
configure it, and do not ask the user for a token.
```
### On the word
`capability` was taken — `CAPABILITIES` is the permission registry. `skill` was taken twice: by
`capabilities/skills/` here and by the wider AI ecosystem.
`affordance` is the exact term for what an environment offers an actor, and the precision earns the
slight awkwardness: rendering an image is not a skill the agent has, it is something the client on the
other end permits. Keeping that distinction in the name saves explaining it every time.
+3 -89
View File
@@ -1,92 +1,6 @@
# Plugins # Plugins
`EXTRACTING-A-PLUGIN.md` is the runbook for turning a platform feature into a plugin, and **The instructions live in [`AGENTS.md`](AGENTS.md). Read it first — this file only points there.**
`EXTRACT-A-PLUGIN-PROMPT.md` is the way to hand that to a fresh agent. This file is for the concepts a
plugin author needs that are not procedure.
**This directory holds documentation, not plugins.** Every plugin is its own repository; `plugins/<name>/` Kept separate so agents that look for `AGENTS.md` by convention find the same guidance as those that
appears by `git clone` when somebody installs one, and is gitignored here. look for `CLAUDE.md`, without the two drifting apart.
---
## Affordances
**`docs/affordances.md` is the authority** — what one is, the three tests for whether something qualifies,
the file format and how they reach the agent. Read it before writing one. What follows is only the part
that is specific to shipping one from a plugin.
> An affordance is something the surface on the other end of the conversation offers the agent — declared
> by the code that implements it, and read by the agent at runtime. Not what the agent can do; what the
> client will honour.
A plugin ships an `affordances/` directory of markdown-with-frontmatter in its repository. Installing
copies it to `capabilities/affordances/<publisher>/<appName>/`, uninstalling removes it — the same shape
as `assets/` and `publishAssets`, and the namespace comes from where it lands rather than from anything
inside the file.
**The test that catches most bad ones: can it be fixed instead?** If the answer is yes, fix it. An
affordance describing a defect gives that defect a rationale, and the rationale outlives everyone's memory
that it was a bug. See the two worked examples in `docs/affordances.md`.
Keep it short — every affordance is loaded into every conversation on this machine, including turns that
have nothing to do with your plugin.
### A plugin-contributed one, for reference## Notifications — the same shape, for push
A plugin can also declare what its push notifications MEAN. Same mechanism, different store:
```
capabilities/notifications/
officerdev/music/
catalogue.yaml
```
```yaml
nowPlaying:
title: 'Now playing'
route: 'officer://music/{id}'
```
That becomes the code `officerdev/music:nowPlaying`. A push carries the code and integers — never content —
and the device renders the title and follows the route. The point is that **adding a notification is a
server change and never an app build**, because the mapping is synced to the device rather than compiled
into it.
**The namespace comes from where the file lands, not from what it says.** Writing `officer/core:test` as a
key in your own catalogue produces `<you>/<yourplugin>:officer/core:test`, not the core code — the same
rule as `mountPrefix`, and pinned by test in `src/servers/notifications/catalogue.test.ts`.
An entry needs a `title`; `body`, `target`, `route` and `thread` are optional, `thread` defaults to your
namespace, and unknown keys are dropped rather than forwarded to the device. Broken YAML in one plugin's
catalogue is skipped without affecting anyone else's.
See `docs/push-notifications.md`.
### A plugin-contributed one, for reference
This is the gitea plugin's, and it shows what these are for: the agent cannot discover from the
filesystem that `tea` is authenticated, or that it should prefer `git` for the things git already does.
```markdown
---
name: Gitea
description: Use when working with git repositories, or with Gitea issues and pull requests.
version: 1
---
For anything git already does — clone, fetch, pull, push, branches, tags — use plain `git`. It talks
to the server directly.
For anything git does not — pull requests, issues, comments, releases, creating a repository — use
`tea`. It is already logged in as the current user, with that user's own permissions. Do not
configure it, and do not ask the user for a token.
```
### On the word
`capability` was taken — `CAPABILITIES` is the permission registry. `skill` was taken twice: by
`capabilities/skills/` here and by the wider AI ecosystem.
`affordance` is the exact term for what an environment offers an actor, and the precision earns the
slight awkwardness: rendering an image is not a skill the agent has, it is something the client on the
other end permits. Keeping that distinction in the name saves explaining it every time.
+1 -1
View File
@@ -89,7 +89,7 @@ async function main(): Promise<void> {
const { PERMISSIONS } = await import('../src/servers/permissions/registry'); const { PERMISSIONS } = await import('../src/servers/permissions/registry');
record( record(
checkCount( checkCount(
join(ROOT, 'CLAUDE.md'), join(ROOT, 'AGENTS.md'),
/([0-9]+|[a-z-]+) entries as of/gi, /([0-9]+|[a-z-]+) entries as of/gi,
PERMISSIONS.length, PERMISSIONS.length,
'permission registry entries', 'permission registry entries',
+271
View File
@@ -0,0 +1,271 @@
# Frontend Apps
Shared patterns and conventions for all frontend applications (dashboard, editor, runtime).
## UI Components
### shadcn/ui (Base)
Location: `src/workspaces/components/ui/`
Standard shadcn/ui components with Tailwind CSS. Import via:
```tsx
import { Button } from "@/components/ui/button";
import { Dialog, DialogContent, DialogTrigger } from "@/components/ui/dialog";
```
### Custom Components
Location: `src/workspaces/components/`
Built on top of shadcn primitives:
- `Avatar.tsx` - User avatars
- `Card.tsx` - Custom card wrapper
- `ColorPicker.tsx` - Color selection
- `DataTable/` - Table with sorting, filtering, pagination (see below)
- `Dialogs/` - Common dialog patterns
- `ErrorDialogs/` - Error display dialogs
- `MetricCard.tsx` - Stats display card
- `SearchInput.tsx` - Search with debounce
- `Select.tsx` - Enhanced select
- `Slider/` - Custom slider
- `Tabs/` - Enhanced tabs
### DataTable + useDataControl
A complete data table solution with sorting, filtering, and pagination.
**Pattern:**
```tsx
import { DataTable, useDataControl } from '@/components/DataTable';
function JobsLibrary() {
const { jobs } = useJobs();
const dataController = useDataControl<Job>(jobs || []);
return (
<>
{/* Wire search input to controller */}
<SearchInput
value={dataController.searchQuery}
handleSearch={dataController.setSearchQuery}
/>
<DataTable<Job>
dataController={dataController}
pageSize={10}
columns={[
{ field: 'id', label: 'ID', sortKey: 'id' },
{ field: 'type', label: 'Type', sortKey: 'type' },
{
field: 'status',
label: 'Status',
condition: view === 'All', // conditional column
format: ({ value, item }) => <Badge>{value}</Badge>
},
{
label: 'Actions',
format: ({ item }) => <ActionsCell job={item} />
},
]}
/>
</>
);
}
```
**useDataControl returns:**
- `data` - Current page of filtered/sorted data
- `rawData` - Original unfiltered data
- `searchQuery`, `setSearchQuery` - Search state
- `searchKeys`, `setSearchKeys` - Which fields to search
- `sortedBy`, `setSortedBy`, `sortBy` - Sort state
- `sortKey`, `sortDirection` - Parsed sort info
- `currentPage`, `changePage`, `pageCount` - Pagination
- `pageSize`, `setPageSize` - Items per page
- `setCustomSort` - Custom sort function
**Column options:**
- `field` - Key in data object
- `label` / `labelMobile` - Header text
- `sortKey` - Enable sorting on this column
- `format` - Custom render: `({ value, item, data, idx }) => ReactNode`
- `condition` - Show/hide column: `boolean | () => boolean`
- `tooltip` - Header tooltip text
- `headerClassName` / `cellClassName` - Styling
**Features:**
- Auto-generates columns from data if not specified
- Resets to page 1 on search/sort change
- Search strips diacritics for accent-insensitive matching
- Built-in pagination bar (shows when data exceeds pageSize)
## Hooks
Location: `src/workspaces/hooks/src/`
### Data Fetching - `useClient`
HTTP client with auth token handling:
```tsx
const client = useClient();
const data = await client.get<User[]>("/users");
await client.post("/experiments", payload);
```
Methods: `get`, `getText`, `getBlob`, `post`, `put`, `patch`, `delete`
Auto-attaches Bearer token from localStorage/sessionStorage.
### Forms - `useForm`
Custom form hook (not react-hook-form):
```tsx
const { state, formRef, update, reset, isValid } = useForm<FormState>(
initialState,
validateFn
);
return <form ref={formRef}>...</form>;
```
Features:
- Auto-syncs form inputs with state via `name` attribute
- Handles checkboxes, radios, selects, number inputs
- Nested object support via data attributes
- Validation function support
### Other Hooks
- `useAuth/` - Authentication state and methods
- `useDebounce` - Debounced values
- `useDragAndDrop` - Drag and drop functionality
- `useFullscreen/` - Fullscreen API wrapper
- `useIsMobile` - Responsive breakpoint detection
- `useLocalStorageState` - Persisted state
- `useMounted` - Component mount status
- `usePopover` - Popover state management
- `useTimeout` - Timeout management
- `useTimer` - Interval-based timer
- `useChatWebSocket` - the chat socket, with reconnect + replay cursor
- `usePanelChannel` - panel-to-panel signals (refresh buses, NOT selection — selection is the URL)
- `useJobs` - background jobs
- `useCustomSorter`, `useImageLoader`, `useIsProduction`, `usePhotoEditor`
## State Management
Uses React Query cache as both server and client state manager.
### Global State - `useGlobal`
Uses React Query cache as a global state store (no Context providers needed):
```tsx
const [value, setValue, refresh, reset] = useGlobal<string>("SIDEBAR_STATE", "expanded");
// Any component using the same key shares state and reacts to changes
setValue("collapsed");
```
How it works:
- `enabled: false` + `staleTime: Infinity` = never fetches, just stores
- Automatic re-renders when state changes
- Visible in React Query DevTools
- Supports functional updates: `setValue(prev => ...)`
### URL Query State - `useQueryState`
Syncs state with URL query parameters:
```tsx
const [page, setPage, reset, clear] = useQueryState<number>("page", 1);
const [filter, setFilter] = useQueryState<string>("filter", null, true); // isGlobal = true
```
Features:
- Updates URL via `history.replaceState` (no page reload)
- `isGlobal = true`: persists across navigation, uses `useGlobal` internally
- `isGlobal = false`: local to component, uses `useState`
- Auto-rebuilds URL when pathname changes (preserves global query state)
- Type coercion based on `defaultValue` type
### Server State
Domain-specific hooks in `state/` directories wrap React Query:
```tsx
// src/workspaces/hooks/src/useJobs.ts
const { jobs, isLoading } = useJobs();
```
### When to Use What
| Scenario | Hook |
|----------|------|
| API data | Domain hooks (`useJobs`, `useModels`, …) |
| Shared UI state | `useGlobal` |
| URL-driven state (filters, pagination) | `useQueryState` |
| Component-only state | `useState` |
| Shared UI state that must survive a refresh, per tab | `useSessionState` |
| Persisted to localStorage | `useLocalStorageState` |
## Event Handlers
**Always use `ev` for event parameters**, not `e`:
```tsx
// ✅ Good
onChange={(ev) => setName(ev.target.value)}
onKeyDown={(ev) => {
if (ev.key === 'Enter') {
ev.preventDefault();
handleSubmit();
}
}}
const handleSubmit = async (ev: React.FormEvent) => {
ev.preventDefault();
// ...
};
// ❌ Bad
onChange={(e) => setName(e.target.value)}
```
## React 19 Patterns
**RefObject includes null**: In React 19, `useRef` returns `RefObject<T | null>`. Update prop types accordingly:
```tsx
// ✅ Good - allow null in ref type
type PanelProps = {
triggerRef: React.RefObject<HTMLButtonElement | null>;
};
// ❌ Bad - will error when passing useRef result
type PanelProps = {
triggerRef: React.RefObject<HTMLButtonElement>;
};
```
**Hook return types**: Use `ReturnType<typeof hookName>` for typing hook returns in props:
```tsx
import { useJobs } from 'hooks/useJobs';
type TopHeaderProps = {
manager: ReturnType<typeof useJobs>;
};
```
## Error Handling
**Toast notifications** via Sonner:
```tsx
import { toast } from "sonner";
toast.error("Something went wrong");
toast.success("Saved successfully");
```
API errors automatically trigger via `useClient.config.onError`.
## Further Reading
- `../../CLAUDE.md` — architecture and the frontend route conventions
- `../../docs/navigation-audit.md` — **authoritative** on routing and selection. The short version:
addressable state lives in the URL (`useParams` / `?selected=`), never in a channel or a global; rows
and nav items are real `<Link>`/`<NavLink>`s. Read it before building a screen that selects things.
- `../../CONVENTIONS.md` — component organisation and React patterns
- `../workspaces/officerdev/APP_CONVENTIONS.md` — panel apps and the AppRegistry
+3 -268
View File
@@ -1,271 +1,6 @@
# Frontend Apps # Frontend Apps
Shared patterns and conventions for all frontend applications (dashboard, editor, runtime). **The instructions live in [`AGENTS.md`](AGENTS.md). Read it first — this file only points there.**
## UI Components Kept separate so agents that look for `AGENTS.md` by convention find the same guidance as those that
look for `CLAUDE.md`, without the two drifting apart.
### shadcn/ui (Base)
Location: `src/workspaces/components/ui/`
Standard shadcn/ui components with Tailwind CSS. Import via:
```tsx
import { Button } from "@/components/ui/button";
import { Dialog, DialogContent, DialogTrigger } from "@/components/ui/dialog";
```
### Custom Components
Location: `src/workspaces/components/`
Built on top of shadcn primitives:
- `Avatar.tsx` - User avatars
- `Card.tsx` - Custom card wrapper
- `ColorPicker.tsx` - Color selection
- `DataTable/` - Table with sorting, filtering, pagination (see below)
- `Dialogs/` - Common dialog patterns
- `ErrorDialogs/` - Error display dialogs
- `MetricCard.tsx` - Stats display card
- `SearchInput.tsx` - Search with debounce
- `Select.tsx` - Enhanced select
- `Slider/` - Custom slider
- `Tabs/` - Enhanced tabs
### DataTable + useDataControl
A complete data table solution with sorting, filtering, and pagination.
**Pattern:**
```tsx
import { DataTable, useDataControl } from '@/components/DataTable';
function JobsLibrary() {
const { jobs } = useJobs();
const dataController = useDataControl<Job>(jobs || []);
return (
<>
{/* Wire search input to controller */}
<SearchInput
value={dataController.searchQuery}
handleSearch={dataController.setSearchQuery}
/>
<DataTable<Job>
dataController={dataController}
pageSize={10}
columns={[
{ field: 'id', label: 'ID', sortKey: 'id' },
{ field: 'type', label: 'Type', sortKey: 'type' },
{
field: 'status',
label: 'Status',
condition: view === 'All', // conditional column
format: ({ value, item }) => <Badge>{value}</Badge>
},
{
label: 'Actions',
format: ({ item }) => <ActionsCell job={item} />
},
]}
/>
</>
);
}
```
**useDataControl returns:**
- `data` - Current page of filtered/sorted data
- `rawData` - Original unfiltered data
- `searchQuery`, `setSearchQuery` - Search state
- `searchKeys`, `setSearchKeys` - Which fields to search
- `sortedBy`, `setSortedBy`, `sortBy` - Sort state
- `sortKey`, `sortDirection` - Parsed sort info
- `currentPage`, `changePage`, `pageCount` - Pagination
- `pageSize`, `setPageSize` - Items per page
- `setCustomSort` - Custom sort function
**Column options:**
- `field` - Key in data object
- `label` / `labelMobile` - Header text
- `sortKey` - Enable sorting on this column
- `format` - Custom render: `({ value, item, data, idx }) => ReactNode`
- `condition` - Show/hide column: `boolean | () => boolean`
- `tooltip` - Header tooltip text
- `headerClassName` / `cellClassName` - Styling
**Features:**
- Auto-generates columns from data if not specified
- Resets to page 1 on search/sort change
- Search strips diacritics for accent-insensitive matching
- Built-in pagination bar (shows when data exceeds pageSize)
## Hooks
Location: `src/workspaces/hooks/src/`
### Data Fetching - `useClient`
HTTP client with auth token handling:
```tsx
const client = useClient();
const data = await client.get<User[]>("/users");
await client.post("/experiments", payload);
```
Methods: `get`, `getText`, `getBlob`, `post`, `put`, `patch`, `delete`
Auto-attaches Bearer token from localStorage/sessionStorage.
### Forms - `useForm`
Custom form hook (not react-hook-form):
```tsx
const { state, formRef, update, reset, isValid } = useForm<FormState>(
initialState,
validateFn
);
return <form ref={formRef}>...</form>;
```
Features:
- Auto-syncs form inputs with state via `name` attribute
- Handles checkboxes, radios, selects, number inputs
- Nested object support via data attributes
- Validation function support
### Other Hooks
- `useAuth/` - Authentication state and methods
- `useDebounce` - Debounced values
- `useDragAndDrop` - Drag and drop functionality
- `useFullscreen/` - Fullscreen API wrapper
- `useIsMobile` - Responsive breakpoint detection
- `useLocalStorageState` - Persisted state
- `useMounted` - Component mount status
- `usePopover` - Popover state management
- `useTimeout` - Timeout management
- `useTimer` - Interval-based timer
- `useChatWebSocket` - the chat socket, with reconnect + replay cursor
- `usePanelChannel` - panel-to-panel signals (refresh buses, NOT selection — selection is the URL)
- `useJobs` - background jobs
- `useCustomSorter`, `useImageLoader`, `useIsProduction`, `usePhotoEditor`
## State Management
Uses React Query cache as both server and client state manager.
### Global State - `useGlobal`
Uses React Query cache as a global state store (no Context providers needed):
```tsx
const [value, setValue, refresh, reset] = useGlobal<string>("SIDEBAR_STATE", "expanded");
// Any component using the same key shares state and reacts to changes
setValue("collapsed");
```
How it works:
- `enabled: false` + `staleTime: Infinity` = never fetches, just stores
- Automatic re-renders when state changes
- Visible in React Query DevTools
- Supports functional updates: `setValue(prev => ...)`
### URL Query State - `useQueryState`
Syncs state with URL query parameters:
```tsx
const [page, setPage, reset, clear] = useQueryState<number>("page", 1);
const [filter, setFilter] = useQueryState<string>("filter", null, true); // isGlobal = true
```
Features:
- Updates URL via `history.replaceState` (no page reload)
- `isGlobal = true`: persists across navigation, uses `useGlobal` internally
- `isGlobal = false`: local to component, uses `useState`
- Auto-rebuilds URL when pathname changes (preserves global query state)
- Type coercion based on `defaultValue` type
### Server State
Domain-specific hooks in `state/` directories wrap React Query:
```tsx
// src/workspaces/hooks/src/useJobs.ts
const { jobs, isLoading } = useJobs();
```
### When to Use What
| Scenario | Hook |
|----------|------|
| API data | Domain hooks (`useJobs`, `useModels`, …) |
| Shared UI state | `useGlobal` |
| URL-driven state (filters, pagination) | `useQueryState` |
| Component-only state | `useState` |
| Shared UI state that must survive a refresh, per tab | `useSessionState` |
| Persisted to localStorage | `useLocalStorageState` |
## Event Handlers
**Always use `ev` for event parameters**, not `e`:
```tsx
// ✅ Good
onChange={(ev) => setName(ev.target.value)}
onKeyDown={(ev) => {
if (ev.key === 'Enter') {
ev.preventDefault();
handleSubmit();
}
}}
const handleSubmit = async (ev: React.FormEvent) => {
ev.preventDefault();
// ...
};
// ❌ Bad
onChange={(e) => setName(e.target.value)}
```
## React 19 Patterns
**RefObject includes null**: In React 19, `useRef` returns `RefObject<T | null>`. Update prop types accordingly:
```tsx
// ✅ Good - allow null in ref type
type PanelProps = {
triggerRef: React.RefObject<HTMLButtonElement | null>;
};
// ❌ Bad - will error when passing useRef result
type PanelProps = {
triggerRef: React.RefObject<HTMLButtonElement>;
};
```
**Hook return types**: Use `ReturnType<typeof hookName>` for typing hook returns in props:
```tsx
import { useJobs } from 'hooks/useJobs';
type TopHeaderProps = {
manager: ReturnType<typeof useJobs>;
};
```
## Error Handling
**Toast notifications** via Sonner:
```tsx
import { toast } from "sonner";
toast.error("Something went wrong");
toast.success("Saved successfully");
```
API errors automatically trigger via `useClient.config.onError`.
## Further Reading
- `../../CLAUDE.md` — architecture and the frontend route conventions
- `../../docs/navigation-audit.md` — **authoritative** on routing and selection. The short version:
addressable state lives in the URL (`useParams` / `?selected=`), never in a channel or a global; rows
and nav items are real `<Link>`/`<NavLink>`s. Read it before building a screen that selects things.
- `../../CONVENTIONS.md` — component organisation and React patterns
- `../workspaces/officerdev/APP_CONVENTIONS.md` — panel apps and the AppRegistry
+231
View File
@@ -0,0 +1,231 @@
# Database Patterns
## Dropping and adding a column in the same push makes drizzle ASK
`drizzle-kit push` cannot tell a drop-plus-add from a rename, so it stops and asks:
```
Is grants_postgres column in roles table created or renamed from another column?
❯ + grants_postgres create column
~ provisioning_script › grants_postgres rename column
```
**Non-interactively that is fatal and silent.** With stdin closed it bails, `bun db:push` still exits 0, and
the `postdb:push` seed then runs against a schema that never changed — which is how you get a confusing
error about a column that does not exist, several steps from the thing that failed.
Hit on 2026-08-23 replacing `roles.provisioning_script` with `grants_postgres`/`grants_docker`. It is the
same family as the composite-unique truncate prompt below, and it matters more on a deploy: no agent driving
a rollout over SSH has a terminal to answer with.
**So a release whose schema drops a column carries the SQL in its plan**, guarded and idempotent, and treats
`bun db:push` as the check that it worked rather than the thing that does it:
```sql
ALTER TABLE roles ADD COLUMN IF NOT EXISTS grants_postgres boolean NOT NULL DEFAULT false;
ALTER TABLE roles ADD COLUMN IF NOT EXISTS grants_docker boolean NOT NULL DEFAULT false;
ALTER TABLE roles DROP COLUMN IF EXISTS provisioning_script;
```
Then push, and expect zero DDL. If it still wants to change something, the SQL was wrong.
## Overview
**One PostgreSQL database**, `officer_db`, exposed as the workspace package `officerdb`. Managed with
Drizzle ORM. There is no second database and no cache database — anything that used to say otherwise was
describing a different codebase.
```
src/databases/officer_db/
├── src/
│ ├── db.ts # the connection
│ ├── index.ts # public surface: one `export * from './<feature>'` per line
│ ├── schema.ts # what db:push creates — see below
│ ├── types.ts # every type export (Select / Insert / extended)
│ ├── crypto.ts # at-rest encryption, one key per purpose
│ ├── secret-store.ts # the key store itself (SQLite, outside Postgres)
│ └── <feature>/
│ ├── index.ts # what this feature exports — declared here, not in a list three levels up
│ ├── schema.ts # its tables
│ └── queries.ts # everything that reads or writes them
└── package.json # exports ".", "./types", "./db", "./schema", "./secret-store", "./*"
```
**A feature owns its own public surface.** `src/index.ts` is one `export *` per feature and nothing
else; what a feature exports is declared in its own `index.ts`, beside the code it describes. Adding a
query function is one file in one directory, rather than that file plus a hand-written list of every
symbol in the package. That list was 297 lines until 2026-08-13 and it had already drifted — twelve
features listed twice, `db` and `schema` buried at line 270 with three feature blocks after them.
**One directory per feature, holding both halves.** Restructured 2026-08-13 from parallel `schema/` and
`queries/` trees, where the two sides had drifted: four features were named differently on each side
(`app-store`/`sidecar-installs`, `email`/`email-accounts`, `server`/`server-config`), `operations` had no
query file at all, and `integrations` had no schema file.
One directory is still lopsided and says so by its contents: `integrations/` has only queries, because it
spans `server` and `user-data`. (`operations/` was the other, and was deleted on 2026-08-13 along with the
Task Logs feature — see below.)
**`src/schema.ts` is drizzle-kit's view, not the runtime's.** `drizzle.config.ts` points at it, so a
commented line there removes a table from the DATABASE without removing a line of code — every query
imports its tables from `./schema` inside its own feature directory. That is what lets a fresh install
create only the core tables, with the plugin ones commented out until their plugin is installed.
## Schema changes use `push`, not migrations
`bun db:push` diffs the schema code against the live database and alters it directly. **`drizzle-kit
migrate` has never been run here** — there is no `__drizzle_migrations` table. Change the schema, run
push, done. `bun db:gen` writes files to `migrations/`, but nothing applies them; treat the schema code
as the source of truth, never those files.
**The schema is the source of truth for what the database may contain**, not just for its shape. Where a
column has a known set of legal values, say so with a `check()` rather than leaving it free `text` —
otherwise a value the code stopped supporting sits there unnoticed. `serverIntegrations.provider` is the
worked example: rows for deleted chat integrations kept their bot tokens long after the code that read
them was gone, because nothing structural said they had become illegal.
```ts
const SERVER_PROVIDERS = ['google', 'apify'] as const;
check(
'ck_server_integrations_provider',
// sql.raw, not sql`${p}` — an interpolated JS string binds as a parameter, so the constraint is
// emitted as `IN ($1, $2)` and Postgres refuses it, breaking push for the WHOLE schema.
sql`${table.provider} IN (${sql.join(
SERVER_PROVIDERS.map((p) => sql.raw(`'${p}'`)),
sql`, `,
)})`,
);
```
Adding a constraint fails while existing rows violate it. That is the point: push refusing tells you the
database has drifted, instead of quietly accepting it. Clean the rows, then push.
### Composite keys: use `uniqueIndex`, and name any long foreign key
**Declare a multi-column uniqueness rule as `uniqueIndex('uq_…').on(a, b)`, never `unique('uq_…').on(a, b)`.**
drizzle-kit mis-diffs named composite unique CONSTRAINTS. It reads them back from the database, compares
them against a schema that declares exactly the same name, columns and order, decides they differ, and
emits a `DROP CONSTRAINT` + `ADD CONSTRAINT` pair — on every push, forever. Fifteen of them made
`db:push` plan 32 statements against a database that already matched, and `ADD UNIQUE` on a populated
table is a data-risk statement, so push stopped on an interactive _"do you want to truncate?"_ prompt
that could never be resolved by answering it.
Reproduced on a database drizzle had itself created seconds earlier, so it is not drift and not
something your change caused. Single-column `.unique()` is diffed correctly and is unaffected — only the
table-level `unique('name').on(...)` form. Unique indexes are diffed on a different code path and are
stable. Verified on drizzle-kit 0.31.9 / drizzle-orm 0.45.1.
A unique index enforces exactly what the constraint did, and `onConflictDoUpdate({ target: [...] })`
accepts it as an arbiter. The one thing it cannot do is be the target of a foreign key — Postgres
requires a unique _constraint_ there. Nothing here has a composite foreign-key target; check before
adding one.
**Name a foreign key explicitly when drizzle's generated name would exceed 63 characters.** Postgres
truncates identifiers at 63 and stores the shortened form, so drizzle keeps comparing against its own
longer version and re-creates the constraint on every push.
`user_integrations_server_integration_id_server_integrations_id_fk` was 65, and is now declared with
`foreignKey({ name: 'fk_user_integrations_server_integration', … })`.
**Known remaining churn, harmless:** a composite `primaryKey` has the same diffing bug and there is no
index form to escape to — a primary key must be a constraint. `music_now_playing` therefore drops and
re-adds `pk_music_now_playing` on every push. Two statements, silent, no prompt even with rows in the
table, data unaffected. Neither naming it nor un-naming it helps — both were tested. Leave it.
**All six over-long foreign keys are now named, 2026-08-23.** They were deferred as "noise rather than a
fault" for four days. What ended that was a deploy plan saying _"expect no changes, stop and report if it
wants to alter something"_ — which this made impossible to satisfy on any host, forcing the deploy agent to
overrule its own instructions and continue. A documented annoyance became a plan instruction nobody could
follow.
The six, and where each lives:
fk_command_hq_watcher_fires_session officer_db/src/command-hq/schema.ts
fk_command_hq_session_tags_session officer_db/src/command-hq/schema.ts
fk_bitcoin_wallet_frozen_utxos_wallet plugins/bitcoin-wallet/db/schema.ts
fk_bitcoin_wallet_chain_cache_wallet plugins/bitcoin-wallet/db/schema.ts
fk_atelier_collection_items_collection plugins/atelier/sidecar/db/collections/schema.ts
fk_atelier_collection_items_generation plugins/atelier/sidecar/db/collections/schema.ts
Two things worth knowing if you add another. Naming it means REMOVING the inline `.references()` on the
column, or drizzle creates both. And `walletChainCache` had to move from the two-argument `pgTable` form to
the three-argument one, because there was nowhere else to put the declaration.
**`pk_music_now_playing` still churns, and it cannot be fixed.** Two statements per push. Tested on
2026-08-23 rather than assumed: removing the explicit name produces the identical drop-and-re-add under
drizzle's generated name (`music_now_playing_user_id_device_pk`), so this is the named-composite mis-diff
rather than the truncation one, and there is no index form to escape to because a primary key must be a
constraint. **A push is therefore expected to print exactly these two statements and nothing else** — which
is a far more useful baseline than ten, because anything else really is your change.
**Rules that still apply:**
1. **Never answer "Yes, truncate the table"** if a prompt ever does appear. It destroys rows and fixes
nothing — whatever is being re-created gets re-created next push regardless.
2. **Never delete a constraint from the schema to quiet a diff.** The schema is right. Removing a
uniqueness rule would break the upserts that depend on it.
3. **Do not use `--force`.** It auto-accepts data-loss statements and nobody has established which
branch it takes at a truncate prompt.
4. **Read the plan before applying it:**
`bunx drizzle-kit push --config=drizzle.config.ts --verbose` prints every statement first. Expect the
two `pk_music_now_playing` lines and the eight from the four over-long foreign keys above — ten in
all. Anything else is your change.
To test a schema change without touching `officer_dev`: `createdb officer_scratch`, then
`POSTGRES_URL=postgresql://postgres:postgres@127.0.0.1:5432/officer_scratch bunx drizzle-kit push
--config=drizzle.config.ts`. Push twice — the second run tells you whether your declaration is stable.
## Type naming
```ts
// types.ts — inferred from the schema, never hand-written
export type UserSelect = typeof Schema.users.$inferSelect;
export type UserInsert = typeof Schema.users.$inferInsert;
// The bare name is the hydrated shape, when a table has relations worth carrying
export type User = UserSelect & {
passkeys: PasskeySelect[];
};
```
Organise `types.ts` by domain with section comments, mirroring the schema files.
## Queries
Hand-written, one `queries.ts` per feature directory, importing tables from `./schema` beside it and
types from `../types`:
```ts
import { eq, and } from 'drizzle-orm';
import { db } from '../db';
import { users, passkeys } from '../schema';
import type { UserSelect } from '../types';
export async function getUsers(): Promise<UserSelect[]> { … }
```
Every query is exported from `src/index.ts`, which is the only surface callers use.
## Importing
```ts
// ✅ types from the types subpath
import type { User } from 'officerdb/types';
// ✅ queries, schema and drizzle helpers from the package root
import { getUserById, eq } from 'officerdb';
// ❌ never reach into src/schema or src/queries directly
// ❌ never hand-write a type a table can infer
```
In app code import from `'types'`, which re-exports the database types. Only server and database code
imports `officerdb/types` directly.
## Conventions
- `serial` for ids on small tables; `bigserial` with `mode: 'number'` where the row count is unbounded
(`chat_session_events` is the example — its id is also the replay cursor).
- `timestamp('…', { withTimezone: true })`, always.
- Index anything you filter on. Drizzle-inferred types reflect nullability correctly, so guard rather
than cast.
+3 -228
View File
@@ -1,231 +1,6 @@
# Database Patterns # Database Patterns
## Dropping and adding a column in the same push makes drizzle ASK **The instructions live in [`AGENTS.md`](AGENTS.md). Read it first — this file only points there.**
`drizzle-kit push` cannot tell a drop-plus-add from a rename, so it stops and asks: Kept separate so agents that look for `AGENTS.md` by convention find the same guidance as those that
look for `CLAUDE.md`, without the two drifting apart.
```
Is grants_postgres column in roles table created or renamed from another column?
❯ + grants_postgres create column
~ provisioning_script › grants_postgres rename column
```
**Non-interactively that is fatal and silent.** With stdin closed it bails, `bun db:push` still exits 0, and
the `postdb:push` seed then runs against a schema that never changed — which is how you get a confusing
error about a column that does not exist, several steps from the thing that failed.
Hit on 2026-08-23 replacing `roles.provisioning_script` with `grants_postgres`/`grants_docker`. It is the
same family as the composite-unique truncate prompt below, and it matters more on a deploy: no agent driving
a rollout over SSH has a terminal to answer with.
**So a release whose schema drops a column carries the SQL in its plan**, guarded and idempotent, and treats
`bun db:push` as the check that it worked rather than the thing that does it:
```sql
ALTER TABLE roles ADD COLUMN IF NOT EXISTS grants_postgres boolean NOT NULL DEFAULT false;
ALTER TABLE roles ADD COLUMN IF NOT EXISTS grants_docker boolean NOT NULL DEFAULT false;
ALTER TABLE roles DROP COLUMN IF EXISTS provisioning_script;
```
Then push, and expect zero DDL. If it still wants to change something, the SQL was wrong.
## Overview
**One PostgreSQL database**, `officer_db`, exposed as the workspace package `officerdb`. Managed with
Drizzle ORM. There is no second database and no cache database — anything that used to say otherwise was
describing a different codebase.
```
src/databases/officer_db/
├── src/
│ ├── db.ts # the connection
│ ├── index.ts # public surface: one `export * from './<feature>'` per line
│ ├── schema.ts # what db:push creates — see below
│ ├── types.ts # every type export (Select / Insert / extended)
│ ├── crypto.ts # at-rest encryption, one key per purpose
│ ├── secret-store.ts # the key store itself (SQLite, outside Postgres)
│ └── <feature>/
│ ├── index.ts # what this feature exports — declared here, not in a list three levels up
│ ├── schema.ts # its tables
│ └── queries.ts # everything that reads or writes them
└── package.json # exports ".", "./types", "./db", "./schema", "./secret-store", "./*"
```
**A feature owns its own public surface.** `src/index.ts` is one `export *` per feature and nothing
else; what a feature exports is declared in its own `index.ts`, beside the code it describes. Adding a
query function is one file in one directory, rather than that file plus a hand-written list of every
symbol in the package. That list was 297 lines until 2026-08-13 and it had already drifted — twelve
features listed twice, `db` and `schema` buried at line 270 with three feature blocks after them.
**One directory per feature, holding both halves.** Restructured 2026-08-13 from parallel `schema/` and
`queries/` trees, where the two sides had drifted: four features were named differently on each side
(`app-store`/`sidecar-installs`, `email`/`email-accounts`, `server`/`server-config`), `operations` had no
query file at all, and `integrations` had no schema file.
One directory is still lopsided and says so by its contents: `integrations/` has only queries, because it
spans `server` and `user-data`. (`operations/` was the other, and was deleted on 2026-08-13 along with the
Task Logs feature — see below.)
**`src/schema.ts` is drizzle-kit's view, not the runtime's.** `drizzle.config.ts` points at it, so a
commented line there removes a table from the DATABASE without removing a line of code — every query
imports its tables from `./schema` inside its own feature directory. That is what lets a fresh install
create only the core tables, with the plugin ones commented out until their plugin is installed.
## Schema changes use `push`, not migrations
`bun db:push` diffs the schema code against the live database and alters it directly. **`drizzle-kit
migrate` has never been run here** — there is no `__drizzle_migrations` table. Change the schema, run
push, done. `bun db:gen` writes files to `migrations/`, but nothing applies them; treat the schema code
as the source of truth, never those files.
**The schema is the source of truth for what the database may contain**, not just for its shape. Where a
column has a known set of legal values, say so with a `check()` rather than leaving it free `text` —
otherwise a value the code stopped supporting sits there unnoticed. `serverIntegrations.provider` is the
worked example: rows for deleted chat integrations kept their bot tokens long after the code that read
them was gone, because nothing structural said they had become illegal.
```ts
const SERVER_PROVIDERS = ['google', 'apify'] as const;
check(
'ck_server_integrations_provider',
// sql.raw, not sql`${p}` — an interpolated JS string binds as a parameter, so the constraint is
// emitted as `IN ($1, $2)` and Postgres refuses it, breaking push for the WHOLE schema.
sql`${table.provider} IN (${sql.join(
SERVER_PROVIDERS.map((p) => sql.raw(`'${p}'`)),
sql`, `,
)})`,
);
```
Adding a constraint fails while existing rows violate it. That is the point: push refusing tells you the
database has drifted, instead of quietly accepting it. Clean the rows, then push.
### Composite keys: use `uniqueIndex`, and name any long foreign key
**Declare a multi-column uniqueness rule as `uniqueIndex('uq_…').on(a, b)`, never `unique('uq_…').on(a, b)`.**
drizzle-kit mis-diffs named composite unique CONSTRAINTS. It reads them back from the database, compares
them against a schema that declares exactly the same name, columns and order, decides they differ, and
emits a `DROP CONSTRAINT` + `ADD CONSTRAINT` pair — on every push, forever. Fifteen of them made
`db:push` plan 32 statements against a database that already matched, and `ADD UNIQUE` on a populated
table is a data-risk statement, so push stopped on an interactive _"do you want to truncate?"_ prompt
that could never be resolved by answering it.
Reproduced on a database drizzle had itself created seconds earlier, so it is not drift and not
something your change caused. Single-column `.unique()` is diffed correctly and is unaffected — only the
table-level `unique('name').on(...)` form. Unique indexes are diffed on a different code path and are
stable. Verified on drizzle-kit 0.31.9 / drizzle-orm 0.45.1.
A unique index enforces exactly what the constraint did, and `onConflictDoUpdate({ target: [...] })`
accepts it as an arbiter. The one thing it cannot do is be the target of a foreign key — Postgres
requires a unique _constraint_ there. Nothing here has a composite foreign-key target; check before
adding one.
**Name a foreign key explicitly when drizzle's generated name would exceed 63 characters.** Postgres
truncates identifiers at 63 and stores the shortened form, so drizzle keeps comparing against its own
longer version and re-creates the constraint on every push.
`user_integrations_server_integration_id_server_integrations_id_fk` was 65, and is now declared with
`foreignKey({ name: 'fk_user_integrations_server_integration', … })`.
**Known remaining churn, harmless:** a composite `primaryKey` has the same diffing bug and there is no
index form to escape to — a primary key must be a constraint. `music_now_playing` therefore drops and
re-adds `pk_music_now_playing` on every push. Two statements, silent, no prompt even with rows in the
table, data unaffected. Neither naming it nor un-naming it helps — both were tested. Leave it.
**All six over-long foreign keys are now named, 2026-08-23.** They were deferred as "noise rather than a
fault" for four days. What ended that was a deploy plan saying _"expect no changes, stop and report if it
wants to alter something"_ — which this made impossible to satisfy on any host, forcing the deploy agent to
overrule its own instructions and continue. A documented annoyance became a plan instruction nobody could
follow.
The six, and where each lives:
fk_command_hq_watcher_fires_session officer_db/src/command-hq/schema.ts
fk_command_hq_session_tags_session officer_db/src/command-hq/schema.ts
fk_bitcoin_wallet_frozen_utxos_wallet plugins/bitcoin-wallet/db/schema.ts
fk_bitcoin_wallet_chain_cache_wallet plugins/bitcoin-wallet/db/schema.ts
fk_atelier_collection_items_collection plugins/atelier/sidecar/db/collections/schema.ts
fk_atelier_collection_items_generation plugins/atelier/sidecar/db/collections/schema.ts
Two things worth knowing if you add another. Naming it means REMOVING the inline `.references()` on the
column, or drizzle creates both. And `walletChainCache` had to move from the two-argument `pgTable` form to
the three-argument one, because there was nowhere else to put the declaration.
**`pk_music_now_playing` still churns, and it cannot be fixed.** Two statements per push. Tested on
2026-08-23 rather than assumed: removing the explicit name produces the identical drop-and-re-add under
drizzle's generated name (`music_now_playing_user_id_device_pk`), so this is the named-composite mis-diff
rather than the truncation one, and there is no index form to escape to because a primary key must be a
constraint. **A push is therefore expected to print exactly these two statements and nothing else** — which
is a far more useful baseline than ten, because anything else really is your change.
**Rules that still apply:**
1. **Never answer "Yes, truncate the table"** if a prompt ever does appear. It destroys rows and fixes
nothing — whatever is being re-created gets re-created next push regardless.
2. **Never delete a constraint from the schema to quiet a diff.** The schema is right. Removing a
uniqueness rule would break the upserts that depend on it.
3. **Do not use `--force`.** It auto-accepts data-loss statements and nobody has established which
branch it takes at a truncate prompt.
4. **Read the plan before applying it:**
`bunx drizzle-kit push --config=drizzle.config.ts --verbose` prints every statement first. Expect the
two `pk_music_now_playing` lines and the eight from the four over-long foreign keys above — ten in
all. Anything else is your change.
To test a schema change without touching `officer_dev`: `createdb officer_scratch`, then
`POSTGRES_URL=postgresql://postgres:postgres@127.0.0.1:5432/officer_scratch bunx drizzle-kit push
--config=drizzle.config.ts`. Push twice — the second run tells you whether your declaration is stable.
## Type naming
```ts
// types.ts — inferred from the schema, never hand-written
export type UserSelect = typeof Schema.users.$inferSelect;
export type UserInsert = typeof Schema.users.$inferInsert;
// The bare name is the hydrated shape, when a table has relations worth carrying
export type User = UserSelect & {
passkeys: PasskeySelect[];
};
```
Organise `types.ts` by domain with section comments, mirroring the schema files.
## Queries
Hand-written, one `queries.ts` per feature directory, importing tables from `./schema` beside it and
types from `../types`:
```ts
import { eq, and } from 'drizzle-orm';
import { db } from '../db';
import { users, passkeys } from '../schema';
import type { UserSelect } from '../types';
export async function getUsers(): Promise<UserSelect[]> { … }
```
Every query is exported from `src/index.ts`, which is the only surface callers use.
## Importing
```ts
// ✅ types from the types subpath
import type { User } from 'officerdb/types';
// ✅ queries, schema and drizzle helpers from the package root
import { getUserById, eq } from 'officerdb';
// ❌ never reach into src/schema or src/queries directly
// ❌ never hand-write a type a table can infer
```
In app code import from `'types'`, which re-exports the database types. Only server and database code
imports `officerdb/types` directly.
## Conventions
- `serial` for ids on small tables; `bigserial` with `mode: 'number'` where the row count is unbounded
(`chat_session_events` is the example — its id is also the replay cursor).
- `timestamp('…', { withTimezone: true })`, always.
- Index anything you filter on. Drizzle-inferred types reflect nullability correctly, so guard rather
than cast.
+118
View File
@@ -0,0 +1,118 @@
# Themes
Drop a theme in here and it appears in the picker on the next page load. No install step, no rebuild, no
restart — a theme is a block of CSS custom properties, and none of that needs compiling.
**This directory is untracked**, exactly like `plugins/`. Only this file is in the repository. Your themes
are yours; the platform's own live in `src/workspaces/theme/css/themes/`.
---
## The shape
```
themes/
my-theme/
theme.css required — the values
theme.json required — the name a person reads
assets/ optional — a wallpaper, a logo
```
The directory name **is** the theme's id. It is not repeated in the manifest, because an id that can
disagree with where its files sit eventually will.
### `theme.json`
```json
{
"name": "My Theme",
"family": "My Theme",
"dark": true
}
```
| | |
|---|---|
| `name` | required — what the picker shows |
| `family` | optional — groups variants together, like `Duck Light` and `Duck Dark`. Defaults to `name` |
| `dark` | optional — adds the `.dark` class. Set it for a dark theme; several hundred `dark:` utilities still resolve against it |
### `theme.css`
Must target its own directory name, or nothing will happen and it will be baffling:
```css
[data-theme='my-theme'] {
--background: #1e1e2e;
--foreground: #cdd6f4;
/* …the rest */
}
```
The server checks this on discovery and reports it rather than letting you wonder.
---
## What to put in it
**Twenty tokens are required.** They are the standard shadcn set, which is what every published palette
already ships — Tokyo Night, Catppuccin, Gruvbox, Nord all come in this vocabulary, so porting one is
mostly copy-paste.
```
background foreground · card card-foreground · popover popover-foreground
primary primary-foreground · secondary secondary-foreground
muted muted-foreground · accent accent-foreground
destructive destructive-foreground · border input ring · radius
```
**Everything else is derived from those.** Panels, the header, the dock, glass, grid lines, scrollbars —
all computed from `card`, `border`, `background` and `foreground`. A palette theme never mentions them.
**Four more are worth setting**, because the standard set has no notion of them and the defaults are
generic: `--success`, `--warning`, `--info`, `--busy`, each with a `-foreground`. `busy` means *running*,
not *warning* — it is what a live job wears.
**Anything else you want to change, you can.** Declaring a variable in your theme beats the derived
default, because both land on `<html>` and yours is more specific. That is how a theme gets a *texture*
rather than just a palette:
```css
[data-theme='my-theme'] {
--panel-bg: rgba(255, 255, 255, 0.08); /* glass instead of a flat card */
--panel-blur: 20px;
--panel-radius: 0; /* square panels */
--page-bg-src: '/themes/my-theme/assets/wallpaper.webp';
--logo-src: '/themes/my-theme/assets/logo.svg';
}
```
The full list of what you can override is in `src/workspaces/theme/README.md`, and
`src/workspaces/theme/css/themes/duck-dark.css` is a worked example of a textured theme.
**Assets are quoted paths, not `url(...)`.** The bundler resolves `url()` at build time and your files are
served at runtime, so a `url()` token would break. The app wraps the path for you.
---
## Writing one
Start from `src/workspaces/theme/css/themes/plain.css` rather than from a Duck theme. Plain is white,
black and grey throughout, so every value you change is one you decided about, with nothing inherited from
someone else's brand.
Then: reload the page, open the picker or press `ctrl+shift+space`, and use `j`/`k` to browse. `l` applies
the focused theme to the app behind the list, so you can watch your own theme against real screens.
Editing `theme.css` and reloading is the whole loop. The stylesheet is cached for a minute.
---
## Checking your work
Render the app under **Plain** first if something looks wrong under yours. Anything still coloured under
Plain is a component painting itself rather than reading a token — that is a platform bug, not yours, and
worth reporting.
`bun run check:theme` counts hardcoded colours in the platform. It does not check themes, but it is why
your theme has a chance of working: the number can only go down.
+3 -115
View File
@@ -1,118 +1,6 @@
# Themes # Themes
Drop a theme in here and it appears in the picker on the next page load. No install step, no rebuild, no **The instructions live in [`AGENTS.md`](AGENTS.md). Read it first — this file only points there.**
restart — a theme is a block of CSS custom properties, and none of that needs compiling.
**This directory is untracked**, exactly like `plugins/`. Only this file is in the repository. Your themes Kept separate so agents that look for `AGENTS.md` by convention find the same guidance as those that
are yours; the platform's own live in `src/workspaces/theme/css/themes/`. look for `CLAUDE.md`, without the two drifting apart.
---
## The shape
```
themes/
my-theme/
theme.css required — the values
theme.json required — the name a person reads
assets/ optional — a wallpaper, a logo
```
The directory name **is** the theme's id. It is not repeated in the manifest, because an id that can
disagree with where its files sit eventually will.
### `theme.json`
```json
{
"name": "My Theme",
"family": "My Theme",
"dark": true
}
```
| | |
|---|---|
| `name` | required — what the picker shows |
| `family` | optional — groups variants together, like `Duck Light` and `Duck Dark`. Defaults to `name` |
| `dark` | optional — adds the `.dark` class. Set it for a dark theme; several hundred `dark:` utilities still resolve against it |
### `theme.css`
Must target its own directory name, or nothing will happen and it will be baffling:
```css
[data-theme='my-theme'] {
--background: #1e1e2e;
--foreground: #cdd6f4;
/* …the rest */
}
```
The server checks this on discovery and reports it rather than letting you wonder.
---
## What to put in it
**Twenty tokens are required.** They are the standard shadcn set, which is what every published palette
already ships — Tokyo Night, Catppuccin, Gruvbox, Nord all come in this vocabulary, so porting one is
mostly copy-paste.
```
background foreground · card card-foreground · popover popover-foreground
primary primary-foreground · secondary secondary-foreground
muted muted-foreground · accent accent-foreground
destructive destructive-foreground · border input ring · radius
```
**Everything else is derived from those.** Panels, the header, the dock, glass, grid lines, scrollbars —
all computed from `card`, `border`, `background` and `foreground`. A palette theme never mentions them.
**Four more are worth setting**, because the standard set has no notion of them and the defaults are
generic: `--success`, `--warning`, `--info`, `--busy`, each with a `-foreground`. `busy` means *running*,
not *warning* — it is what a live job wears.
**Anything else you want to change, you can.** Declaring a variable in your theme beats the derived
default, because both land on `<html>` and yours is more specific. That is how a theme gets a *texture*
rather than just a palette:
```css
[data-theme='my-theme'] {
--panel-bg: rgba(255, 255, 255, 0.08); /* glass instead of a flat card */
--panel-blur: 20px;
--panel-radius: 0; /* square panels */
--page-bg-src: '/themes/my-theme/assets/wallpaper.webp';
--logo-src: '/themes/my-theme/assets/logo.svg';
}
```
The full list of what you can override is in `src/workspaces/theme/README.md`, and
`src/workspaces/theme/css/themes/duck-dark.css` is a worked example of a textured theme.
**Assets are quoted paths, not `url(...)`.** The bundler resolves `url()` at build time and your files are
served at runtime, so a `url()` token would break. The app wraps the path for you.
---
## Writing one
Start from `src/workspaces/theme/css/themes/plain.css` rather than from a Duck theme. Plain is white,
black and grey throughout, so every value you change is one you decided about, with nothing inherited from
someone else's brand.
Then: reload the page, open the picker or press `ctrl+shift+space`, and use `j`/`k` to browse. `l` applies
the focused theme to the app behind the list, so you can watch your own theme against real screens.
Editing `theme.css` and reloading is the whole loop. The stylesheet is cached for a minute.
---
## Checking your work
Render the app under **Plain** first if something looks wrong under yours. Anything still coloured under
Plain is a component painting itself rather than reading a token — that is a platform bug, not yours, and
worth reporting.
`bun run check:theme` counts hardcoded colours in the platform. It does not check themes, but it is why
your theme has a chance of working: the number can only go down.