Downstream packages lost the tests that exercised their agent-loop-facing
edges when the loop was rewritten. Restore 100% per-file coverage with
behavior tests through public seams: llm-retry config validation and
cancellation races, goal replay drift/staleness/teardown edges, plan-mode
disposed-flush, workspace-context empty-change commits, api-proxy
synchronous send failures, acp-snapshot spill-path extraction and refresh
write-back, ACP injection-triggered turns, cli-demo and tui inbox
lifecycle edges, and agent-loop retry/settlement/lifecycle branches.
The only source changes are narrowly-justified v8 ignore annotations on
invariant guards and one redundant-guard removal (workspace-context).
The branch edited the English side of 17 doc pairs without their Chinese
counterparts. Bring every .zh.md up to date (loop.ts references now point
at agent.ts, send/steer/inject unification, context/message folded into
user/message, HookContext replaced by UserMessageData in tool decision
blocks, plan/skill delivery moved to agent/step) and re-record the pairing
hashes. Deduplicate type-equiv.manifest.json entries duplicated by the
master merge and drop entries for symbols the refactor removed. Complete
the missing JSDoc on extractSnapshotSpillPaths.
This branch intentionally appends plugin-sourced context (plan-mode
notices, the skill catalog) after the prompt, so the request's last message
is no longer the user's text. The scripted adapter keyed every scenario off
messages.at(-1) only, missed its triggers, and fell through to the
ask_user_question default — the frozen 'Waiting for the first token' PTY
timeouts. It now scans all text since the last assistant message.
Prompt admission runs before any turn opens, so the agent/prompt-submit
flush listener could only ever append plan/mode outside an open turn —
invariant-vetoed and retried at agent/step when the session invariant is
mounted, silently out-of-turn in the durable log when it is not. agent/step
fires inside the open turn before every request derivation (including
turn 1 step 1), so it is the sole flush point and no behavior is lost: a
pending idle switch lands as the first step's in-turn plan/mode. Tests pin
the new contract: admission never appends; the first step boundary flushes.
concludesTurn is terminal: a completed step now carries the concluded fact
to the driver, which ends the turn immediately instead of letting steering
that arrived during the tool batch (or from an agent/stopping listener)
reopen it. The steering is not lost — it drained into the log as
steering/message before the decision point, so it feeds the next turn's
request. Also annotates the driver's unreachable invariant guards
(exhaustiveness default, abort-slot ownership) for the coverage gate.
A pre-commit turn/start rejection previously left the machine bricked:
lastTurn had already advanced past a turn the log never recorded, so every
later turn/start violated the session invariant's contiguity rule, and the
admitted prompt lingered in the outbox to leak into the next turn's request.
Keep admitted input on the stack (an argument to run()) until turn/start
commits, then advance lastTurn, set turnOpen, and append the prompt and its
additional contexts as user/message events inside the now-existing turn.
A rejected turn/start therefore unwinds with zero shared state to roll
back, the turn number stays reusable, and the outbox never holds input for
a turn that does not exist. This also restores the documented event order:
the prompt follows turn/start directly instead of waiting in the outbox
behind any steering carried over by cancel({keepInbox}).
macOS symlinks /tmp → /private/tmp; when the session header cwd lacks
the /private prefix but fs tools resolve symlinks, tool result paths
get the prefix and the cwd replacement never matches. Strip the prefix
before cwd→{{cwd}} replacement so both forms normalize identically.
Also extend refreshFixtureReplacements to match spill paths by filename
suffix, preventing raw fixture churn when a refresh runs from a
different session directory.
The fake terminal retains earlier autocomplete frames, so the directory follow-up assertion could pass before the scoped lookup completed. Wait for the quoted file mention to be applied before submitting.
_Kerman
Tianyi Cui
Turtle