Add BenchDeck local taste skills#14
Conversation
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: a2ef7bde73
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".
| - Use `benchdeck-terminal-taste` for TUI layout, hierarchy, legibility, and narrow-width polish. | ||
| - Use `benchdeck-screenshot-quality` for screenshot candidate generation, review, and visual-evidence handoffs. | ||
| - Use `benchdeck-readme-polish` for README/docs presentation, screenshot captions, and product-story clarity. | ||
| - Use `benchdeck-output-completeness` when a task requires complete plans, files, audits, handoffs, or validation reports. |
There was a problem hiding this comment.
This new root rule applies benchdeck-output-completeness to audit tasks, but the preceding audit rule sends audits to @repo-auditor and that agent still has skill: deny in .opencode/agents/repo-auditor.md. Because this same file tells agents to stop when requirements conflict, a repository audit that needs a complete handoff or validation report has no way to satisfy both instructions. Route the skill through repo-auditor or remove audits from this rule.
Useful? React with 👍 / 👎.
| 3. **Build evidence ledger.** For each material claim, record path, line or section, evidence rating, source type, and contradiction status. | ||
| 4. **Detect stale content.** Compare docs against source/tests/config. Severity: high for direct contradiction/broken command; medium for incomplete current behavior; low for wording/age signals. | ||
| 5. **Classify edit risk.** Apply low-risk fixes directly when allowed: broken relative links, verified command names, stale file paths, typos, missing cross-links, and clarified wording. Ask before changing promises, install/release/security guidance, feature status, examples with side effects, or broad rewrites. | ||
| 5. **Classify edit risk.** Apply low-risk fixes directly when allowed: broken relative links, verified command names, stale file paths, typos, missing cross-links, and clarified wording. Ask before changing promises, install/release guidance, feature status, examples with side effects, or broad rewrites. |
There was a problem hiding this comment.
For /docs-update tasks that touch security-sensitive docs under the allowed docs/** edit scope, such as runner token/PAT guidance, this line no longer requires approval before changing security guidance; the previous wording did. That means security documentation changes can now be classified as ordinary low-risk wording/path fixes, so please keep security guidance in the explicit ask-before-changing list.
Useful? React with 👍 / 👎.
- Route benchdeck-output-completeness skill through repo-auditor to resolve conflict with AGENTS.md audit rule - Restore security guidance to the approval-gate list in repository-docs.md so security docs changes require explicit approval
1 participant
Summary
.opencode/skills/adapted from the taste-skill implementation plan.benchdeck-terminal-taste,benchdeck-screenshot-quality,benchdeck-readme-polish, andbenchdeck-output-completeness..opencodeallowlists fortui-precision-editor,tui-screenshot, andrepository-docs.AGENTS.mdso future agents know when to use each local skill.Safety
Validation
main...agent/taste-skill-integration; changed paths are limited to.opencode/agents/*,.opencode/skills/*, andAGENTS.md.Known limitation
repo-auditoragent was not changed. An attempted routing update for that file was blocked by the connector, so the skill index was corrected to list only agents that were actually routed.