Repository navigation
Pin vendor skills for CC lane messaging, evals and logs - #852
Conversation
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
Dependency limit exceeded — report not shown. This pull request scan exceeded the 10,000-dependency limit applied to this scan, so the results are incomplete and may be inaccurate. To avoid reporting false positives, Socket has not posted a report. Upgrade your plan to raise the dependency limit and get complete reports, or view the partial scan in the dashboard. Socket is always free for open source. If this is a non-commercial open source project, contact us to request a free Team account. |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 0a243fcadf
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
1694009 to
ab5280c
Compare
Scope
Pin the maintainers'
hcom-agent-messaging,promptfoo-evalsandlokiskills for command-center and native lane work. Their descriptions add 890 characters; the repository counter reports Claude 10,374/10,500 and Codex 10,055. Keeppromqldeferred to the owner's cap decision and keep the runtime-worker selection at 134 skills with explicit exclusions for these three lane picks.The decision compares each vendor workflow with native documentation/operation routes and maintained community definitions. Promptfoo's skill uses the supported profile's 0.123.1 source revision; its entire folder is identical at 0.124.0, so the correction preserves the instructions and listing totals.
28af9b9321d0492f6cdabe75091376560e7cdc4d.lane:foundation.When
lokiappears in the native ClaudeskillOverridesmap, the renderer's service-name check treats that key as a skill identifier. Actual Loki endpoints, MCP server references and override values remain checked for missing service wiring. The scan does not change the rendered configuration or declare Loki wired.SOTA sources
v0.7.28:skills/hcom-agent-messaging/SKILL.mdandLICENSE.0.123.1:.claude/skills/promptfoo-evals/SKILL.md, its references andLICENSE. Whole folder tree5c0e65e6c185b6a1ea6212626b8de5442193ed76is identical at0.124.0.skills/grafana-lgtm/loki/SKILL.mdandLICENSE.adoption/skills/lifecycle.md.jsonmodule.Evidence-class table
source_reviewlocal_integrationscripts/skills_status.py:budget_reportlocal_integrationlocal_integrationlocal_integrationpython3 scripts/validate.py: exit 0Local commands run
The existing renderer regression and endpoint/MCP/value controls also pass (four focused native methods). The decision retains failed conditions from initial local and hosted validation, including the source/task projections and the skill/service namespace collision. Marketplace scan verdicts remain Unknown; their links are locators.
Decision record
docs/decisions/2026-10-08-skills-cc-lanes-picks.mdrecords the scoped source comparison, supported Promptfoo revision, counters, source-review correction, worker exclusions, inverse and per-selection overturn conditions. Applicability is source evidence; comparative outcomes and client invocation acceptance remain unmeasured.Host evidence
Installation and invocation are deferred to the designated host shell outside Claude after this PR lands (CC plan row 31). This PR records no new skill use, evaluation, component install or live client settings change. The paper-runtime freeze and quiet window remain binding.
Checklist