Refresh the generated capability catalog from command-code@1.44.0 to
1.53.1. deepseek/deepseek-v4.1-flash now advertises image input and its
low, high, and max reasoning efforts, so pi forwards attached images and
exposes the full thinking-level selector instead of reporting the model
as text-only without selectable levels. The sync also adds gpt-6-astra,
xai/grok-4.6, and inclusionai/ling-3.0-flash-sante:free metadata.
Drop the now-obsolete manual Meta Muse Spark effort overrides because
upstream publishes efforts for meta/muse-spark-1.1 through 1.3-contributor
and MiniMaxAI/MiniMax-M3, and add a regression test for the V4.1
capabilities.
Muse Spark 1.3 and 1.3-contributor are discoverable via the Provider API
but Pi showed "Current model does not support thinking" because the
generated catalog had no reasoning flag or selectable efforts.
Reuse the existing Muse Spark manual override mechanism:
- add reasoning flags and image modalities for both ids to the generated
catalog (CLI 1.40.1 snapshot follows the same pattern as 1.1/1.2)
- add a manual effort override of [minimal, low, medium, high, xhigh]
in src/commandcode-catalog-overrides.ts, merged over the generated
catalog at load time (same set as 1.1/1.2/1.2-contributor, no max)
Pi selected thinking levels now map to reasoning_effort via the
existing generic thinkingMetadataForModel / mappedReasoningEffort path;
no model-specific transport logic is needed.
Tests: typecheck, test-models (override merge + thinking metadata),
test-model-metadata-check, test-pure-functions, test-pricing, test-cost,
test-stream (reasoning_effort mapping), test-runtime, test-transport all pass.
No live Command Code API test was performed.
Co-authored-by: heie54 <14918606+heie54@user.noreply.gitee.com>
Co-authored-by: Patrick Wozniak <email@patwoz.de>
Oh My Pi kept the unresolved $COMMAND_CODE_API_KEY placeholder as a literal config API key that shadowed its /login credential store and was sent as the Bearer token (401). The placeholder is now registered only on pi, where it keeps the API-key auth method and --api-key working next to OAuth; on OMP the provider omits apiKey unless a real key is configured. Host-supplied placeholders are resolved or stripped on every stream path, and the legacy generate transport uses the same rule.
Stored /login OAuth and API-key credentials, --api-key, and env keys are now covered end to end on both pi and Oh My Pi, and CI runs the pi suite against a real binary.
Co-authored-by: ebreen <ebreen@users.noreply.github.com>
Bump the memory benchmark runtime pins to the current stable releases:
Bun 1.3.11 -> 1.4.0, Node 22.19.0 -> 22.23.2, pi host 0.82.1 -> 0.84.4.
Pin the Oh My Pi compatibility job to the same Bun release instead of
'latest' so a required check does not drift on its own.
tests/test-omp-compat.mjs skipped whenever omp was not on PATH, which
was always the case in CI, so 0.6.1 shipped an import that pi's compat
entrypoint provides but OMP's bundled pi-ai does not (#74).
Add an omp-compat job that installs @oh-my-pi/pi-coding-agent with Bun
and runs the suite with OMP_COMPAT_REQUIRED=1, which turns the skip into
a failure. Add a first phase that loads the extension through
`omp models -e`, the same loader `omp plugin install` validates with;
it fails on the 0.6.1 index.ts and passes with the runtime resolution.
0.6.1 imported registerApiProvider from @earendil-works/pi-ai/compat as
a named import. Oh My Pi maps that specifier onto its own pi-ai, which
has no such export and instead registers custom APIs inside
registerProvider, so omp plugin install rejected the extension with
"Export named 'registerApiProvider' not found".
Import the compat module as a namespace and call registerApiProvider
only when the host provides it. pi keeps the sibling-extension fix from
#68; omp loads again and its own custom-api registry covers that case.
Verified with omp 18.1.2: plugin install of the linked checkout succeeds
and tests/test-omp-compat.mjs passes; tests/test-pi-local.mjs still
covers the compat registry path on pi.
Closes#74
The standalone section read like a commit message and sat between Install
and Authentication as if it were a core topic. Keep the part users need
(sibling extensions share the connection and credentials, so their
requests count against Command Code usage) as one sentence under Usage;
the registry mechanics stay in index.ts and the changelog.
google/gemini-3.7-flash currently fails on the Provider API with a
zero-data-retention routing 404 for every request, which made the GOAT
live suite red regardless of provider changes. GPT-5.6 Luna is available
on every plan, accepts image input, and completes the vision request.
PR #69 added efforts for meta/muse-spark-* directly to the generated
catalog, which the drift check flags and the daily sync job reverts.
Keep src/commandcode-catalog.ts byte-identical to upstream and merge a
separate src/commandcode-catalog-overrides.ts over it at load time. A
test fails as soon as upstream publishes efforts for an overridden model
so the override gets removed instead of shadowing the CLI catalog.
Verified against the live endpoint: pi --thinking xhigh sends
reasoning_effort="xhigh" for meta/muse-spark-1.2-contributor and the
request succeeds; --thinking off sends none.
Closes#69
The official CLI marks meta/muse-spark-* as reasoning models but ships no
effort levels for them and sends no effort parameter, so the thinking level
stays off in pi. Document that so the locked footer is not mistaken for a
provider bug.
Refs #69
(cherry picked from commit 263310635a89e35cf0b3da527474555c9687903e)
Every host start awaited a full catalog request before the provider
registered, costing one HTTPS round-trip (up to the discovery timeout on a
hanging connection) even when a cache written seconds earlier was on disk.
Register the cached catalog immediately and run the live refresh in the
background; the live result re-registers the provider when it arrives. A
first start without a cache still awaits the live catalog. The background
refresh is aborted on session_shutdown so print mode does not wait for it.
Closes#63
(cherry picked from commit 142191f9420fe90460cdc3cbae70e26d0c000860)
Declare the fixture's pi types inline so npm run typecheck passes without
the optional peer packages, matching the existing fixture convention.
(cherry picked from commit 0ba5827c58a0f6ccf54387454ae38d93bbead3d9)
pi routes the main chat through the registered provider, but sibling
extensions that call streamSimple from @earendil-works/pi-ai/compat with
the active Command Code model resolve model.api through the compat
api-registry, which only knows built-in APIs. On plain pi that failed
with "No API provider registered for api: commandcode-custom".
Register commandcode-custom there and delegate to the transport router.
The registry resolves no credentials for extension providers, so fall
back to the configured Command Code key when the caller passes none.
Closes#68
(cherry picked from commit 7e9659e672771c6a9223b95938e50fb2a051a0a7)
Verify display pricing against the official pricing page and the live
provider model list:
- add claude-fable-5-1 and deepseek/deepseek-v4-flash-fast
- remove minimax/minimax-m3-free and minimax/minimax-m2.7-free, which the
provider no longer lists
- end the claude-sonnet-5 introductory window; the listed rate is unchanged
- google/gemini-3.7-flash returns to list price after the promotion ended
Add claude-fable-5-1, deepseek/deepseek-v4-flash-fast, and
tencent/hy4-preview with their reasoning efforts, add moonshotai/Kimi-K3
efforts, and drop minimax/minimax-m3-free, which the CLI no longer lists.
The model catalog tests hard-coded specific model ids, the reasoning model
count, and output limits from command-code@1.32.2. Every upstream catalog
sync broke them, which made the daily catalog sync workflow fail before it
could open its PR.
Assert structural invariants over the generated catalog instead: image
models resolve to text+image, every effort entry has a reasoning flag,
reasoning without efforts yields an empty level map, output limits are
positive integers and clamp to the context length.
Closes#67
(cherry picked from commit 458e3a57892bb625a78fb5ded092627ba3cf2867)
Maps the three Muse Spark models to [minimal, low, medium, high, xhigh]
so pi-ai can surface graded thinking levels instead of off-only.
These ids are already MODEL_REASONING=true; without MODEL_EFFORTS,
thinkingMetadataForModel() returns an empty map and the footer stays
:off with no cycle via /thinking or Shift+Tab.
Upstream command-code@1.32.2 lists these models as Efforts: — (none);
this commit intentionally adds a pi-side policy mapping aligned with
the opencode/opencode-go provider catalogs (which ship
muse-spark@ minimal..xhigh). The catalog header notes it is normally
generated by sync:commandcode-catalog — happy to move this to a
separate override layer if maintainers prefer.
Add an agent skill that walks through a full model catalog refresh: drift
detection, catalog sync, manually reviewed pricing updates, fixture
refresh, and test updates. Its helper scripts run on Windows and Linux:
one snapshots the live model-id list into the test fixture, the other
regenerates the pricing fixture from MODEL_COSTS through prettier so
format:check stays green. Typecheck now covers the skill scripts.
execFile cannot spawn npm's .cmd shim directly on Windows, so the catalog
sync and drift check failed with spawn npm ENOENT. Route npm invocations
through the shell with argument quoting on Windows and keep direct
execFile calls elsewhere.
Add the free minimax/minimax-m3-free and minimax/minimax-m2.7-free
promotional variants (free through September 5, 2026) and
tencent/hy4-preview, drop the retired stealth/ox-alpha, and refresh the
model-id and pricing snapshots to the current 62-model catalog verified
on 2026-08-28.
Qwen/Qwen3.8-Flash lists $0.16 input, $0.47 output, and $0.016 cache read
per million tokens, and z-ai/glm-5.3-flash lists $0.15, $0.50, and $0.03;
the official pricing page documents no cache-write rate for either model.
Synchronize the static model catalog with command-code@1.36.0, which adds
Qwen/Qwen3.8-Flash (reasoning efforts low, medium, xhigh) and
z-ai/glm-5.3-flash (low, high, max), the free minimax/minimax-m3-free
model, and the glm-5.3-flash output limit, and drops the retired
stealth/ox-alpha.
Align the developer-message fix with the Provider API transport router.
The fix stays scoped to the legacy /alpha/generate converter, and the
OMP advisory compat test now forces the upgrade_required fallback so the
advisory is asserted on the generate request body.
cat-shark
