The backend injects a single managed model key (THT_MODEL_API_KEY[_FILE]) as
the selected provider's env var, but that key belongs to one provider — so
selecting a second cloud provider (e.g. DeepSeek while the managed key is zai's)
forced the wrong key onto it and failed auth. This is why the model could not be
switched to DeepSeek.
When the selected provider is present in pi's own auth store
(~/.pi/agent/auth.json), skip injection and let pi resolve that provider's key
itself. Deployments without an auth store (containers) yield an empty set, so the
managed-key injection stays authoritative and fail-fast there. authProviders is
injectable into PiProcessManager for deterministic tests.
Verified live: GLM 5.2, DeepSeek V4 Flash, and aritmolab Qwen3.6 all operate through the ThothII model selector.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The FE derived 'working' purely as activeSession && !pendingWidget, so the
final workflow turn — the only one that ends without a follow-up gate —
left the spinner on forever (observed live: 21592s after F8 approve).
- SessionBridge maps Pi's agent_end -> SSE system_event {event: agent_end}
- PiProcessManager notifies the client (info error + synthetic agent_end)
when the child dies unexpectedly; expected teardowns stay silent
- sessionStore tracks agentActive (on: user entry/text_delta/ui_request,
off: agent_end); AppShell working now requires it; resume sets it
optimistically
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
ctx.ui.input in `pi --mode rpc` correlates extension_ui_response on its own
top-level RPC id (crypto.randomUUID), not the descriptor id the gate carries
in `title`. SessionBridge replied with the descriptor id, so Pi silently
dropped the response and the model never resumed — every reviewer widget hung
after the human answered.
SessionBridge now stores Pi's top-level m.id (pendingPiId) and replies
extension_ui_response{ id: pendingPiId, value: <uiResponse> }; value still
carries the descriptor id so the gate's internal resp.id === descriptor.id
check still holds.
The fake-pi double had masked the bug by forcing m.id == descriptor.id; it now
mirrors real Pi (distinct randomUUID, correlate on it, drop unknown ids), with
a negative regression test. SKILL.md Phase 1 also now steers multi-answer
disambiguation to reviewer_decide (multiselect).
Tests: backend 67/67, tsc clean, fake-pi contract 2/2.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
pi 0.73 rpc mode is headless and runs tools without an approval gate; the
removed --approve flag made pi exit with 'Unknown option: --approve', breaking
every session spawn. Spawn args are now just --mode rpc. +regression test.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Found via live browser test: PUT /settings preflight was rejected because
@fastify/cors default methods omit PUT. GET/POST were unaffected.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
workspace/provider/model/thinking now come from getSettings() injected into
sessionRoutes; the request body supplies only question+name. Also teaches
fake_pi_rpc to respond to set_model and set_thinking_level RPC commands so
tests that pass real model settings don't hang.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
marcopanandClaude Opus 4.8