master
4
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
  pastilhasandClaude Opus 5
|
d7f3223b6a |
qwen-local: the first provider that runs on our own GPU
sd-server (stable-diffusion.cpp, Vulkan) serving Qwen-Image 2.1 on this machine. It already speaks OpenAI's image API, which is the vocabulary `ImageRequest` is written in, so there is no translation layer — only the parts that are true of a local GPU and not of a hosted one. ── keyless is a category, not an exception ── `credentials.ts` decided a provider was usable by asking whether a key existed, so a service that takes none read as unconfigured forever. The fix is an adapter declaring `keyless`, not a dummy secret: a dummy would record a credential that does not exist and make the secret store lie about what is configured, and that store is the one place meant to be the truth about them. A keyless adapter never reaches `keyFor` at all. That function's own comment says why it uses `getSecret` rather than `getKey` — so a missing FOREIGN credential is never minted — and a provider with no credential by design has no business in that path. whisper, OCR and TTS already run on this machine, so this will not be the last. ── measured, not assumed ── The integration doc said "tens of seconds, not two". The server's own log says 512x512 completes in ~185s and 1024x1024 samples at ~41.4s/step over 20 steps, about 14 minutes. And it serialises strictly: a second caller waits for the first to finish entirely, so a queued 512 can wait 17 minutes before a byte comes back. So: 512 is the default and the substitute for `auto`, 1024 is offered with the cost in its label, and `timeout: false` is a requirement rather than caution — every default anyone would reach for is far under 17 minutes. The caller's `signal` still aborts, because a user cancelling is not a client giving up. ── the model id moved, because only it could ── `GET /v1/models` returns `sd-cpp-local`, hardcoded in sd-server with no override. The catalogue now agrees with the server and carries "Qwen-Image 2.1 (local)" in the label, so requests name a model the server knows while the UI still says something meaningful. ── not shipped: edit() ── Same GPU, same queue. It doubles the ways generation can be blocked while the generate path is unproven in anger, and `ImageAdapter` makes absence the way to say so — the route answers 400 instead of the queue finding out late. Verified: /v1/models answers, provider statuses report qwen-local as configured/keyless/usable, and the shipped entry survives the merge with the stored catalogue.json. NOT verified end to end — a generation is occupying the GPU and I left it alone. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Officer-Session: de3ed4e6-f88a-429c-94e6-a760336ba83b |
||
|
pastilhas
|
19e94129a0 |
refactor(providers): a provider is a vendor and a key, and the table is gone
The dropdown was listing "connections" — a row per adapter + model + key — which made choosing a model an act of configuration, put a key icon and a bin inside a menu item, and showed five accounts nobody had created. A provider is now a vendor Officer curates. Every one is listed, set up or not, because a vendor you have no key for is exactly what you need to see in order to add one. The model is chosen per generation, from that provider's entry in the catalogue, beside it. With the model off the row, all that remained of a "provider" was the key for OpenAI — and Officer already stores keys: one active value per purpose, 0600 inside a 0700 directory, outside the backup tarball, created on first use so nothing needs provisioning. So `atelier_providers` is dropped, along with `generations.provider_id` that pointed into it. `provider` and `model` stay as text on each generation, which were always the historical fact rather than a pointer to configuration. `getSecret`, never `getKey`: this is a foreign credential, and getKey mints a random value for a purpose it cannot find — which for someone else's API key means authenticating with something invented. Resolution follows: a request names a model, and the catalogue says whose it is, rather than a database being asked which stored row served it. A model nobody in the catalogue claims is still accepted — the catalogue is advisory, and a model released today should work before we ship a file that mentions it. Two routes replace five: PUT and DELETE on /providers/:name/key. Deleting a key that came from the environment is refused rather than faked — it belongs to the install, not to this screen. |
||
|
pastilhas
|
3c59bf4d0c |
feat(catalogue): refresh is a button, not a schedule
The daily job is gone. A refresh only matters at one moment — you are picking a provider and want to know its models are current — and at that moment a button answers the question, where a nightly pass can only have already answered it, invisibly, or not. It also bought a standing obligation: something hitting a vendor from every install forever whether anyone opens Atelier or not. So the add dialog is now the curated file made visible. Provider comes from the catalogue, models come from that provider's entry in it, and two buttons re-read it from the vendor: one for the provider in front of you, one for all of them. `POST /providers/catalogue/refresh?provider=` is the only trigger there is. The dialog also says how old the answer is — "as shipped" until a refresh has ever run, which is worth showing, because it means the models listed are exactly what the plugin was released with. A provider with no refresher answers 400 rather than a silent success: from a button, "nothing happened" and "nothing could happen" look identical. |
||
|
pastilhas
|
0de38f0b13 |
feat(catalogue): a curated file of providers and models, refreshed daily
The five connections in the database were never configured by anyone — a migration minted one per model from the adapter's own list, so the table held five decisions nobody made. What that list actually is, is a CATALOGUE: what you can choose from, which is not the same kind of thing as what you chose. So it is a file. `sidecar/catalogue/shipped.ts` is the curated baseline, in git, reviewed; `$OFFICER_ROOT/data/atelier/catalogue.json` is what a refresh writes, and it wins where it has something to say. Data dir, not the plugin dir — a plugin directory is a clone an update replaces, which is the rule env.ts already states for images. Refresh runs in the sidecar, daily, and only on boot when the file is actually stale — this process has restarted twenty times in an afternoon and should not hit OpenAI twenty times for it. Two sources, unequal, and the code says so. `GET /v1/models` is authoritative and thin: ids and nothing else, so a model matching /^(gpt-image|dall-e)/ appears the day it ships without a release. Prices come off an HTML page because there is no pricing endpoint for any plan, and that parser is written to give up rather than guess — a guess gets written to disk and then believed. When it gives up it says so in the log; the durable replacement is handing the page to a model for structured output, which Officer can already do locally. Nothing a refresh cannot establish overwrites a curated value: fields merge one by one, not objects wholesale, so an id-only listing cannot wipe the sizes and prices that were written by hand. Retiring is decided by the merge, which can tell a complete listing from a request that failed halfway — and it marks `deprecated` rather than deleting, because the gallery labels 88 historical rows from this list. Per-model sizes, qualities and transparency now travel with each entry, which is what will let the form stop offering dall-e-2 a transparent background it cannot do. The catalogue is advisory throughout: validate.ts still refuses to enforce per-model rules, exactly as before. |