Agent skill
install-skill
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/install-skill
SKILL.md
Install Skill
Install skills using local repos (../skills + ../hooks) first. Do not require npm publish.
Primary Flow
- Ensure target skill exists under
../skills/<skill-name>/SKILL.md. - If hooks contracts changed, rebuild local packages:
cd ../hooks
npm run build -w @4meta5/skill-loader
npm run build -w @4meta5/skills
npm run build -w @4meta5/skills-cli
- Install into target project from local source:
cd ../skills
node ../skillex/packages/skills-cli/bin/skills.js add <skill-name> --cwd /absolute/path/to/project
- Sync project skill references:
cd /absolute/path/to/project
node ../skillex/packages/skills-cli/bin/skills.js claudemd sync
Dual Runtime Requirement (Claude + Codex)
The install is not complete until both runtimes are covered:
- Claude Code path:
.claude/skills/<skill-name>/SKILL.mdexistsCLAUDE.mdcontains- @.claude/skills/<skill-name>/SKILL.md.claude/hooks/skill-forced-eval.shexists and points to local CLI fallback when available
- Codex path:
AGENTS.mdmust expose the same installed skill references.- Preferred pattern:
AGENTS.mdsymlink toCLAUDE.md. - If symlink is not used, keep
AGENTS.mdskill refs in lockstep withCLAUDE.md.
Verification commands:
cd /absolute/path/to/project
ls -l AGENTS.md
rg -n "@.claude/skills/.*/SKILL.md" CLAUDE.md AGENTS.md
Hard-Cut Rename Install
If skill names were renamed, do not keep aliases.
- Install new name.
- Remove old name from project:
cd ../skills
node ../skillex/packages/skills-cli/bin/skills.js remove <old-skill-name> --cwd /absolute/path/to/project
- Run
node ../skillex/packages/skills-cli/bin/skills.js claudemd sync --cwd /absolute/path/to/project. - Re-check
AGENTS.mdmirrors updated references.
Replace Set Workflow
When user requests a strict target set (replace stale installs completely):
# 1) remove everything not in target set
cd /absolute/path/to/project
ls .claude/skills
# remove non-target skills explicitly via skills.js remove
# 2) install target set from local skills repo
cd ../skills
node ../skillex/packages/skills-cli/bin/skills.js add <skill-a> <skill-b> ... --cwd /absolute/path/to/project
# 3) sync docs and verify both runtimes
node ../skillex/packages/skills-cli/bin/skills.js claudemd sync --cwd /absolute/path/to/project
cd /absolute/path/to/project
rg -n "@.claude/skills/.*/SKILL.md" CLAUDE.md AGENTS.md
Safety Guardrails For Bulk Cleanup
Before removing skills from .claude/skills:
- Never treat dot-prefixed entries as skills (
.git,.claude,.scout, etc.). - If
.claude/skillsis a symlink, resolve it and verify it points to a dedicated skills directory. - Only remove names that:
- are in a known stale-name list, or
- are installed skill directories with a valid
SKILL.md.
- After cleanup, verify repository metadata still exists (
.gitpresent) and rerunskills list.
Auto-Maintenance
After any incident, update this skill and ../make-skill with:
- Symptom
- Root cause
- Detection command
- Minimal fix command
When invoked, always prefer local-path installation before recommending npm-published binaries.
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?