3.8 KiB
3.8 KiB
Skill Registry
Delegator use only. Any agent that launches sub-agents reads this registry to resolve compact rules, then injects them directly into sub-agent prompts. Sub-agents do NOT read this registry or individual SKILL.md files.
See _shared/skill-resolver.md for the full resolution protocol.
User Skills
| Trigger | Skill | Path |
|---|---|---|
| When creating a pull request, opening a PR, or preparing changes for review | branch-pr | /Users/freetlab/.config/opencode/skills/branch-pr/SKILL.md |
| When creating a GitHub issue, reporting a bug, or requesting a feature | issue-creation | /Users/freetlab/.config/opencode/skills/issue-creation/SKILL.md |
| When user says "judgment day", "judgment-day", "review adversarial", "dual review", "doble review", "juzgar", "que lo juzguen" | judgment-day | /Users/freetlab/.config/opencode/skills/judgment-day/SKILL.md |
| When user asks to create a new skill, add agent instructions, or document patterns for AI | skill-creator | /Users/freetlab/.config/opencode/skills/skill-creator/SKILL.md |
Compact Rules
Pre-digested rules per skill. Delegators copy matching blocks into sub-agent prompts as ## Project Standards (auto-resolved).
branch-pr
- Every PR MUST link an approved issue — no exceptions
- Every PR MUST have exactly one
type:*label - Automated checks must pass before merge is possible
- Branch names must match:
^(feat|fix|chore|docs|style|refactor|perf|test|build|ci|revert)/[a-z0-9._-]+$ - Conventional commits:
^(build|chore|ci|docs|feat|fix|perf|refactor|revert|style|test)(\([a-z0-9\._-]+\))?!?: .+ - Commit type determines PR label: feat→type:feature, fix→type:bug, docs→type:docs, refactor→type:refactor, chore→type:chore, style→type:chore, perf→type:feature, test→type:chore, build→type:chore, ci→type:chore, revert→type:bug
- PR body must contain: Closes #N (linked issue), PR type checkbox, Summary, Changes Table, Test Plan
issue-creation
- Blank issues are disabled — MUST use a template (bug report or feature request)
- Every issue gets
status:needs-reviewautomatically on creation - A maintainer MUST add
status:approvedbefore any PR can be opened - Questions go to Discussions, not issues
- Bug report template required fields: Pre-flight Checks, Bug Description, Steps to Reproduce, Expected Behavior, Actual Behavior, Operating System, Agent/Client, Shell
- Feature request template required fields: Pre-flight Checks, Problem Description, Proposed Solution, Affected Area
judgment-day
- Launch TWO sub-agents via delegate (async, parallel — never sequential)
- Each agent receives the same target but works independently
- Neither agent knows about the other — no cross-contamination
- Classify warnings as WARNING (real) or WARNING (theoretical)
- If confirmed CRITICALs or real WARNINGs exist → delegate Fix Agent
- After Fix Agent completes → re-launch both judges in parallel
- After 2 fix iterations, if issues remain → escalate to user
skill-creator
- Create a skill when: pattern is used repeatedly, project-specific conventions differ, complex workflows need steps, decision trees help AI
- Don't create a skill when: documentation exists, pattern is trivial, one-off task
- Skill structure: frontmatter (name, description, triggers, allowed-tools), Critical Rules, When to Use, Patterns, Commands
Project Conventions
| File | Path | Notes |
|---|---|---|
| SPEC.md | /Users/freetlab/Proyectos/farolero/SPEC.md | Existing SDD artifacts with Explore/Propose/Spec/Tasks/Apply/Verify phases |
| .gga | /Users/freetlab/Proyectos/farolero/.gga | Gentleman Guardian Angel config (AI provider, file patterns, rules file) |
Read the convention files listed above for project-specific patterns and rules. All referenced paths have been extracted — no need to read index files to discover more.