Repository navigation
Conversation
Move the canonical skills to .agents/skills/ (Agent Skills layout read natively by Codex) and expose each one to Claude Code through a relative per-skill symlink in .claude/skills/. Cut the set to new-domain, design-system and sync-template: - new-domain ends with the backend self-review from domain-reviewer - design-system absorbs add-shadcn-component and the ui-reviewer pass - verify and release become always-on rules in AGENTS.md - the Claude-only domain-reviewer and ui-reviewer subagents are removed scripts/tests/test_agent_skills.py checks spec frontmatter, links and the symlink for every skill. Docs, ownership map and init output follow. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
One skills folder for Claude Code and OpenAI Codex, cut down to the must-haves.
.agents/skills/<name>/SKILL.md(Agent Skills spec; Codex reads it natively). Claude Code only discovers.claude/skills/, so each skill there is a relative symlink.claude/skills/<name> -> ../../.agents/skills/<name>(per-skill links are documented; a symlinked skills directory is not). Frontmatter is onlyname+description.new-domain— scaffolder workflow; the formerdomain-reviewerchecklist is its self-review step.design-system— tokens, App* vocabulary, Russian copy; now also adding shadcn-vue primitives (formeradd-shadcn-component) and the browser review (formerui-reviewer).sync-template— fork upstream sync.verifyandreleasebecame always-on rules inAGENTS.md(make checkbefore declaring done; commit type → release, never tag or bump by hand)..claude/agents/removed (Claude-only duplicates of the checklists now inside the skills). Claude Code hooks stay.scripts/tests/test_agent_skills.pychecks frontmatter, links and symlinks; CI runs the tooling tests when.agents/**or.claude/skills/**change.core.symlinks=true(or WSL); otherwise only Codex sees the skills (noted in README).Sources: Codex skills, Claude Code skills, Agent Skills spec.
Test plan
pytest ../scripts/tests: 75 passed, 1 skipped (hook test needs the frontend installed)git ls-files -s .claude/skills— mode 120000 for all three links;SKILL.mdreadable through them🤖 Generated with Claude Code