This release keeps dependencies and maintenance posture current for teams operating this tool.
Published 1mo
AI Agents & Assistants
✓ No known CVEs patched
✓ No known CVEs patched in this version
Topics
ai-agents
aider
anthropic
autonomous
ci-cd
claude
+13 more
cline
code-review
devops
gemini
github-action
github-issues
loki-mode
multi-agent
openai-codex
openapi
pull-request-review
sdlc
spec-driven-development
Summary
AI summaryBench command clarifies missing harness as a development feature and test harness now extracts both load_state and _loki_state_file.
Full changelog
Post-release hardening (UX + test integrity)
- loki bench now degrades honestly on a packaged install. The benchmark harness
(~8MB, mostly historical result dumps) is intentionally not shipped in the npm
package; cmd_bench previously emitted a bare "harness not found" that read like
a broken install. It now explains it is a development/research feature and
points to running it from a source checkout (rc 0, not a failure). - Test integrity: the state-baseline-lifecycle test harness extracted load_state
but not its _loki_state_file dependency (added with the v7.79.0 multi-build
state isolation), so the harness silently skipped the load/reset block and went
red in CI while the runtime was correct. The harness now extracts both
functions. No runtime behavior change (verified: load_state resets
ITERATION_COUNT to 0 on a terminal status exactly as before). - CI parity: local-ci cherry-picked individual shell suites instead of mirroring
the CI "Shell tests" job (tests/run-all-tests.sh), so the harness break above
passed the local gate and only failed post-push. local-ci now mirrors the
state-baseline, loki-why, and bench suites so this class is caught pre-push.
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
About Loki Mode
Multi-agent autonomous SDLC framework. Spec to deployed app. PRD, GitHub issue, OpenAPI/JSON/YAML, or one-line brief. 5 AI providers, 11 quality gates.
Related context
Related tools
Beta — feedback welcome: [email protected]