← Back to Skills Marketplace
austindixson

Launchagent Manager

by austindixson · GitHub ↗ · v1.0.0
cross-platform ✓ Security Clean
328
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install launchagent-manager
Description
List, classify, prune LaunchAgents; analyze openclaw.json so the proper gateway LaunchAgent remains connected and tokens match. Keeps only OpenClaw-related a...
README (SKILL.md)

Launchagent Manager

Description

List, classify, prune LaunchAgents; analyze openclaw.json so the proper gateway LaunchAgent remains connected and tokens match. Keeps only OpenClaw-related agents; can unload/delete others.

LaunchAgent Manager

Manages LaunchAgents in ~/Library/LaunchAgents and analyzes openclaw.json so the gateway LaunchAgent stays correct: loaded when it should be, and tokens matching config vs running gateway.

  • List/prune: Classify agents as OpenClaw (keep) or other (prune). OpenClaw = Label or path contains "openclaw".
  • Config check: Read openclaw.json → gateway port, auth mode, token set; find the gateway plist (e.g. ai.openclaw.gateway); report loaded? running? tokens match? (Uses gateway-guard status when available.) Recommends loading the plist or running gateway-guard ensure --apply if needed.
  • --fix: With --config, can load the gateway plist if not loaded and run gateway-guard ensure --apply if tokens mismatch.

Usage

  • Ensure gateway stays connected: Run --config to verify the gateway LaunchAgent is loaded and tokens match; use --config --fix to load plist and sync auth.
  • List what's running: see OpenClaw vs other agents.
  • Remove non-OpenClaw LaunchAgents: prune so only OpenClaw daemons remain.
python3 \x3Cskill-dir>/scripts/launchagent_manager.py [--list] [--json]
python3 \x3Cskill-dir>/scripts/launchagent_manager.py --config [--fix] [--json]
python3 \x3Cskill-dir>/scripts/launchagent_manager.py --prune [--dry-run]
python3 \x3Cskill-dir>/scripts/launchagent_manager.py --prune --apply [--delete-plists]
  • --list (default) — List all LaunchAgents; show OpenClaw (kept) vs others (prune targets). Shows loaded/unloaded.
  • --config — Analyze openclaw.json and gateway LaunchAgent: config path, gateway port, auth, token set; gateway plist label and loaded?; gateway process running?; tokens match (config vs running)? Recommendations if plist not loaded or tokens mismatch. Exit 0 if all ok, 1 if action needed.
  • --config --fix — If gateway plist not loaded: run launchctl load \x3Cplist>. If tokens don't match: run gateway-guard ensure --apply --json. Requires gateway-guard skill.
  • --config --json — Machine-readable report: config_path, gateway, gateway_launchagent, gateway_loaded, tokens_match, gateway_running, recommendations.
  • --json — For --list: { "openclaw": [...], "others": [...] }.
  • --prune — Operate on non-OpenClaw agents. Without --apply this is a dry-run (show what would be unloaded).
  • --prune --dry-run — Only show what would be unloaded.
  • --prune --apply — Unload each non-OpenClaw LaunchAgent. Plist files kept unless --delete-plists.
  • --prune --apply --delete-plists — Unload and delete plist files (backed up to OPENCLAW_HOME/backups/launchagents).

Safety

  • Only user domain is touched: ~/Library/LaunchAgents/. System domain is not modified.
  • OpenClaw detection is conservative: Label com.openclaw.* or any ProgramArgument containing "openclaw" → kept.
  • With --delete-plists, backups are written to OPENCLAW_HOME/backups/launchagents/ before deletion.

Requirements

  • macOS (launchctl, plist in user LaunchAgents).
  • Python 3 with plistlib (standard library).
Usage Guidance
This skill appears coherent with its description, but it can stop and delete your user LaunchAgents. Before using destructive flags: 1) run with --prune --dry-run to see what would be unloaded/deleted; 2) verify OPENCLAW_HOME points where you expect (defaults to ~/.openclaw) so backups go to the right place; 3) ensure openclaw.json is valid and reviewed (it may contain tokens) and that gateway-guard is the intended local script if the tool invokes it; 4) review the list output before running --prune --apply or --prune --apply --delete-plists. If you are unsure, avoid --apply/--delete-plists or keep a manual backup of ~/Library/LaunchAgents first.
Capability Analysis
Type: OpenClaw Skill Name: launchagent-manager Version: 1.0.0 The skill 'launchagent-manager' is designed to manage macOS LaunchAgents, specifically focusing on OpenClaw's gateway agent and pruning other user-level agents. The Python script `scripts/launchagent_manager.py` uses standard macOS `launchctl` commands to list, load, and unload LaunchAgents within the user's `~/Library/LaunchAgents` directory. It reads `openclaw.json` to verify gateway configuration and interacts with the `gateway-guard` skill for status and fixes. All actions, including file deletion (with backups to `OPENCLAW_HOME/backups/launchagents`), are explicitly stated in the documentation (`SKILL.md`, `README.md`) and are directly aligned with the skill's purpose. There is no evidence of unauthorized data exfiltration, malicious execution, persistence outside its stated management role, or prompt injection attempts in the markdown files. The code's behavior is transparent and justified by its functionality.
Capability Assessment
✓ Purpose & Capability
Name/description match the implementation: the script scans ~/Library/LaunchAgents, parses plists, reads openclaw.json and can load/unload plists and invoke a local gateway-guard script. It does not request unrelated credentials, binaries, or system-wide config.
ℹ Instruction Scope
SKILL.md and the script legitimately read openclaw.json (including token presence) and local plist files and call launchctl and an optional local gateway-guard script. This is expected for its purpose, but the tool can be destructive (--prune --apply --delete-plists) and will read token-related fields from openclaw.json; run with --dry-run first and review backups before deletion.
✓ Install Mechanism
No install spec; the skill is instruction/script-only. No remote downloads or archive extraction are present, lowering install risk.
✓ Credentials
The skill requires no credentials or declared env vars. It optionally respects OPENCLAW_HOME (default ~/.openclaw) for config and backups, which is proportional to its function. It does not attempt to exfiltrate data or call external endpoints.
✓ Persistence & Privilege
always:false and user-invocable. The script operates in the user LaunchAgents domain only and writes backups under OPENCLAW_HOME; it does not modify other skills or request permanent elevated presence.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install launchagent-manager
  3. After installation, invoke the skill by name or use /launchagent-manager
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of Launchagent Manager. - Lists, classifies, and prunes LaunchAgents in ~/Library/LaunchAgents, keeping only OpenClaw-related agents. - Analyzes openclaw.json to ensure the correct gateway LaunchAgent is connected and tokens match. - Offers --config and --fix options to verify and correct LaunchAgent status and token sync. - Supports prune operations to safely unload or delete non-OpenClaw agents, with dry run and backup options. - Provides both human-readable and machine-readable (--json) output formats for reporting and automation.
Metadata
Slug launchagent-manager
Version 1.0.0
License —
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Launchagent Manager?

List, classify, prune LaunchAgents; analyze openclaw.json so the proper gateway LaunchAgent remains connected and tokens match. Keeps only OpenClaw-related a... It is an AI Agent Skill for Claude Code / OpenClaw, with 328 downloads so far.

How do I install Launchagent Manager?

Run "/install launchagent-manager" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Launchagent Manager free?

Yes, Launchagent Manager is completely free (open-source). You can download, install and use it at no cost.

Which platforms does Launchagent Manager support?

Launchagent Manager is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Launchagent Manager?

It is built and maintained by austindixson (@austindixson); the current version is v1.0.0.

More Skills
self-improving agent
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
⬇ 463283 · by pskoett
Skill Vetter
Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
⬇ 259410 · by spclaudehome
Polymarket
Query Polymarket prediction markets. Check odds, find trending markets, search events, track price movements.
⬇ 232503 · by joelchance
Self-Improving + Proactive Agent
Self-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...
⬇ 200423 · by ivangdavila
Github
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
⬇ 191115 · by steipete
ontology
Typed knowledge graph for structured agent memory and composable skills. Use when creating/querying entities (Person, Project, Task, Event, Document), linkin...
⬇ 190156 · by oswalpalash

💬 Comments