Menu
Manage synchronization workflows between repository skills and local system skill directories.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
| name | sync-skills-manager |
| description | Manage synchronization workflows between repository skills and local system skill directories. |
Manage synchronization between repository skills and local/system skill directories.
sync-skills.sh (canonical manager)Sync between repository-canonical categories and runtime installs (~/.claude/skills + other agents).
| Command | Description |
|---|---|
./sync-skills.sh diff | Preview system-only skills |
./sync-skills.sh pull | Sync ~/.claude/skills -> repo (add new skills) |
./sync-skills.sh push | Sync repo -> ~/.claude/skills and refresh agent links |
./sync-skills.sh link-all | Rebuild other agent dirs as symlinks to ~/.claude/skills |
./sync-skills.sh dedupe | Remove duplicate entries from ~/.gemini/skills |
./sync-skills.sh status | Show sync status |
sync-skills-3way.sh (recommended)Incremental 3-way sync across:
~/.codex/skills (including .system)~/.claude/skills (canonical)Key behavior:
rsync --update), never deletes files.tools-skills/ by default (or legacy system-skills/tools-skills/ if present).SKILL.md copy is treated as canonical for repo -> local sync.| Command | Description |
|---|---|
./sync-skills-3way.sh sync | Run 3-way incremental sync (default) |
./sync-skills-3way.sh status | Show counts and name-level diffs |
./sync-skills-3way.sh help | Show usage |
scripts/skills_profiles.py (profiles manager)Manage which skills are enabled per agent by creating/removing symlinks that point to the canonical registry (~/.claude/skills).
This solves two common issues:
Config
system-skills/sync-skills-manager/skills-profiles.jsonSafety model
apply and normalize default to dry-run. Use --apply to change files.SKILL.md.SKILL.md, e.g. dist/) are treated as unmanaged and never touched.normalize always moves existing entries into a backup folder before replacing them with symlinks.Commands
# Show registry + per-agent status
python3 system-skills/sync-skills-manager/scripts/skills_profiles.py status
# Preview what would change (desired vs current)
python3 system-skills/sync-skills-manager/scripts/skills_profiles.py diff
# 1) Sync Codex/registry/repo (incremental, no delete)
python3 system-skills/sync-skills-manager/scripts/skills_profiles.py sync
# 2) Normalize drift (copies/non-canonical links -> canonical symlinks, with backups)
python3 system-skills/sync-skills-manager/scripts/skills_profiles.py normalize --dry-run
python3 system-skills/sync-skills-manager/scripts/skills_profiles.py normalize --apply
# 3) Apply enable/disable sets (remove extra canonical links, add missing links)
python3 system-skills/sync-skills-manager/scripts/skills_profiles.py apply --dry-run
python3 system-skills/sync-skills-manager/scripts/skills_profiles.py apply --apply
# One-shot daily workflow (sync -> normalize -> apply)
python3 system-skills/sync-skills-manager/scripts/skills_profiles.py refresh --apply
# Manage by star ratings (1-7)
# Keep only 7-star skills (other agents). claude-code registry itself is skipped.
python3 system-skills/sync-skills-manager/scripts/skills_profiles.py stars --mode only --stars 7 --dry-run
python3 system-skills/sync-skills-manager/scripts/skills_profiles.py stars --mode only --stars 7 --apply
# Batch install by stars (e.g. add 6-star and 5-star)
python3 system-skills/sync-skills-manager/scripts/skills_profiles.py stars --mode install --stars 6,5 --apply
# Batch uninstall by stars
python3 system-skills/sync-skills-manager/scripts/skills_profiles.py stars --mode uninstall --stars 5 --apply
Backups
~/.claude/skills-backups/<timestamp>/<agent>/<skill>/...Tests
python3 -m unittest -q
scripts/agent_skills_audit.py (agent installation + diff auditor)Check which supported agents are installed locally, list recognized skills, and diff skill sets between agents.
# Scan all supported agents
python3 system-skills/sync-skills-manager/scripts/agent_skills_audit.py scan
# Only installed agents, with skills preview
python3 system-skills/sync-skills-manager/scripts/agent_skills_audit.py scan --installed-only --with-skills
# List full skills for selected agents
python3 system-skills/sync-skills-manager/scripts/agent_skills_audit.py skills --installed-only --agent codex,cursor
# Diff codex vs cursor
python3 system-skills/sync-skills-manager/scripts/agent_skills_audit.py diff --left codex --right cursor
# Diff one baseline against multiple agents
python3 system-skills/sync-skills-manager/scripts/agent_skills_audit.py diff --left codex --right cursor,amp,claude-code
# Sync check: compare installed agents against canonical claude-code
python3 system-skills/sync-skills-manager/scripts/agent_skills_audit.py sync-check --canonical-agent claude-code --installed-only
# Sync check for selected agents
python3 system-skills/sync-skills-manager/scripts/agent_skills_audit.py sync-check --canonical-agent claude-code --agent codex,cursor,amp
scripts/reclassify_system_skills.py (category rebalancer)Rebalance system-skills category distribution (defaults to reclassifying tools-skills into domain categories).
# Preview reclassification plan
python3 system-skills/sync-skills-manager/scripts/reclassify_system_skills.py --from-category tools-skills
# Apply reclassification
python3 system-skills/sync-skills-manager/scripts/reclassify_system_skills.py --from-category tools-skills --apply
scripts/flatten_system_skills_layout.py (layout simplifier)Flatten legacy system-skills/<category>/<skill> into repo-root <category>/<skill> while preserving system-skills/sync-skills-manager/.
# Preview flatten/migration actions
python3 system-skills/sync-skills-manager/scripts/flatten_system_skills_layout.py
# Apply migration actions
python3 system-skills/sync-skills-manager/scripts/flatten_system_skills_layout.py --apply
./sync-skills-3way.sh sync
./sync-skills-3way.sh status
sync-skills.sh uses sync-config.json:
{
"system_skills_path": "~/.claude/skills",
"repo_skills_path": "./",
"exclude_patterns": ["sync-skills-manager"],
"sync_mode": "incremental",
"default_command": "diff"
}
Plan and run Codex-native dynamic workflows for complex tasks that benefit from explicit orchestration, goal mode, subagents or simulated work packets, approval gates, integration, verification, and reusable workflow artifacts. Use when the user invokes this skill, asks for dynamic workflows, subagents, parallel agents, swarm-like work, Goal Maker orchestration, large audits, migrations, multi-track research plus implementation, or Claude Code-style fan-out/fan-in workflows.
Direct browser control via CDP. Use when the user wants to automate, scrape, test, or interact with web pages. Connects to the user's already-running Chrome.
This skill should be used for browser automation tasks using Chrome DevTools Protocol (CDP). Triggers when users need to launch Chrome with remote debugging, navigate pages, execute JavaScript in browser context, capture screenshots, or interactively select DOM elements. No MCP server required.
Use Orca's computer-use CLI to inspect and control local desktop apps through accessibility trees, screenshots, and safe UI actions. Use when an agent needs to list desktop apps, get an app state, read visible UI, click, type, press keys, scroll, drag, set values, or perform app accessibility actions. Triggers include "computer use", "orca computer", "list apps", "get app state", "read Spotify", "read Slack", "click app", "type text", "press key", "set value", "scroll app", "drag app", and desktop app interaction tasks.
Use the `orca` CLI to drive a running Orca editor — manage Orca worktrees; create and manage scheduled automations; create, read, and run shell commands in Orca-managed terminals; and automate Orca's built-in browser (snapshot/click/fill/screenshot/tabs). Use this instead of raw `git worktree`, ad hoc shell PTYs, or Playwright whenever the task touches Orca state. Coding agents inside an Orca worktree should also use it to keep the worktree comment fresh at meaningful checkpoints. Boundary with `orchestration`: if the recipient of a terminal write is another AI agent (Claude Code, Gemini, Codex, a worker), use `orchestration` — it is the only correct way to send messages, nudges, replies, or task hand-offs to agents. orca-cli writes are for non-agent terminals (shells, build/test commands); reading or `wait`ing on any terminal — including agent terminals — stays in orca-cli.
Use Orca orchestration for structured multi-agent coordination: threaded messages, blocking ask/reply flows, task dispatch, worker_done/escalation waits, task DAGs, decision gates, coordinator loops, or decomposing work across agents. Use `orca-cli` instead for ordinary terminal control, lightweight terminal prompts, shell commands, Orca worktree management, reading or waiting on terminals, and automation of the browser embedded inside Orca. Use Computer Use for browser windows, webviews, Orca app UI, or desktop UI outside Orca's embedded browser.