test: green the unit suite + add blocking CI unit-test job#461
Merged
Conversation
…unit-test job
40 -> 0 failures. Root-cause fixes (code or test), never masked.
code:
- pi_only transcript persistence: pi_only drives the argent loop, which discarded
a non-ArgentSessionManager and wrote turns to a throwaway file. Open
ArgentSessionManager.open(sessionFile) so the default runtime mode persists.
- session-fork: write the branch header when absent (pi 0.74 createBranchedSession
flushes lazily).
- intent-validation: createRequire(".js") -> static ESM import (only resolved in dist/).
- skills: stamp `source` back onto loaded skills (bundled-allowlist peekaboo leak).
- TUI select lists: selectCancel -> Key.escape (pi 0.74 action-id rename).
- channel-config tool: Union -> Array (drop forbidden anyOf in tool schema).
- pi-embedded sanitize: gate tool-result synthesis on allowSyntheticToolResults.
- 8 service-key tools: sync resolveServiceKey -> async resolveServiceKeyAsync.
- keychain: short-circuit shell-outs under VITEST (kills the macOS "Keychain Not
Found" modal that stole focus on every test run).
test (stale expectations vs deliberately-advanced source):
- telegram address-prefix, cooldown formula, app-forge view kinds, load-profile,
onboarding (TUI hatch moved to installer), web-tools de-US locale, doctor mock.
- model-auth ENOTEMPTY cleanup flake -> fs.rm maxRetries.
ci/infra:
- add `unit-test` task to the checks matrix (pnpm test:unit).
- quarantine executive-shadow-client.integration (spawns the Rust argent-execd daemon,
absent in CI).
- skip the memory-recall kernel-continuity test (explicit non-goal).
tsgo 189 -> 187. No new failures.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A literal `~/.argentos/memory.db{,-shm,-wal}` folder was swept into the repo in
#436 (a `~` that never expanded into $HOME). These are NOT the real personal
memory DB (~/.argentos/memory.db) — just churning throwaway WAL state that showed
as perpetually-modified and would pollute every "commit everything". Remove from
tracking (working copies already parked) and gitignore the literal `/~/` folder
so they can't be re-committed.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|
Important Review skippedAuto reviews are disabled on base/target branches other than the default branch. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Repository UI Review profile: CHILL Plan: Pro Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
…atch candidates The `build` job runs tsc-since (plain tsc), which flagged 7 net-new errors tsgo accepted: the upstream pi `Skill` type declares no `source` property (it's stamped at runtime by loadSkillEntries), and the `satisfies` + Boolean() predicate pair broke narrowing. Widen toSkillSourceLabel to take `skill: unknown` with an explicit cast, and return an explicitly-typed SkillMatchCandidate with a `!== null` predicate. tsc-since: 7 net-new → 0 (189 baseline). Skills suite: 72 passed. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01T3dkQzz2mSrSNtGBE9cmXr
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Drives
vitest.unit.config.tsfrom 39–40 failed → 0 failed (7806 passed, exit 0) and adds a blockingunit-testjob to thechecksmatrix so a red unit suite can never ship silently again. tsgo 189 → 187. No new failures.Root-cause fixes (never masked)
code:
ArgentSessionManagerand wrote turns to a throwaway file. Now opensArgentSessionManager.open(sessionFile).createBranchedSession); intent-validationcreateRequire→ESM; skillssource-stamp (peekaboo allowlist leak);selectCancel→Key.escape; channel-configanyOf→array; tool-result synthesis gate; 8 service-key tools sync→asyncresolveServiceKeyAsync.VITEST— kills the macOS "Keychain Not Found — ArgentOS" modal that stole focus on every test run.test (stale expectations vs deliberately-advanced source): telegram address-prefix, cooldown formula, app-forge view kinds, load-profile, onboarding (TUI hatch moved to installer), web-tools
de-USlocale, doctor mock, model-authENOTEMPTYflake (fs.rm maxRetries).ci/infra:
unit-testtask (pnpm test:unit); quarantineexecutive-shadow-client.integration(spawns Rustargent-execd, absent in CI);it.skipthe memory-recall kernel-continuity test (explicit non-goal).Also
chore: untracks an accidentally-committed~/.argentos/memory.db*throwaway folder (a tilde that never expanded, swept in perf(#405): take the personal-skill block off the turn critical path (writes + reads) #436) + gitignores/~/.Verify
🤖 Generated with Claude Code