agentleFS
NobiResearchNobiResearch/Ovenbird/.claude/skills/startup/SKILL.md

startup

Skill1 starsChanged 17 days agoMITView on GitHub

Explain what starts with this Mac (login items, launch agents and daemons, background items) and what's worth a look. Use when the human asks what runs at startup or login on their Mac, why the Mac is slow to start, about background items, or whether something on their Mac is suspicious. Not for a project's or app's own startup code.

13 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 9 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Never call something malware without evidence.

    line 34

“Common” is measured: kept by at least 6 in 10 of the 9 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Works from any folder on this Mac (0009). Run syscalls as cos …; in the Ovenbird folder, bin/cos … is the same. Whatever you cos remember or cos friction add is about the human or this Mac, never the project or client you're working in.

Startup

Steps

  1. cos startup --json. It's read-only and needs no permission: the launch agents and daemons (the launchd plists, with their on/off state), plus who signed each program. It doesn't list apps that open at login. If the human wants those too, tell them first that macOS will show an admin password dialog, then run cos startup --full --json (it asks for an Allow). Never run --full on your own or in a headless run.
  2. Lead with the answer: how many things start, and whether anything needs attention.
  3. Items worth a look come first (findings). For each one, say in plain language what it probably is (use the name, developer and path; you know most common apps and tools), why it was flagged, and what you'd do:
    • program missing: left behind by something that was removed; safe to remove the item.
    • unsigned / ad hoc: often a developer tool or something the human built (Homebrew builds are ad-hoc signed); only worrying if they don't recognize it.
    • runs a script: scripts can't be signed; read what it runs (the path) and say so.
    • on but unused for months: a candidate to turn off.
  4. Then a short summary of the rest, grouped (apps, launch agents, daemons). No wall of items.
  5. Changing things: point to System Settings › General › Login Items & Extensions. Removing a user launch agent (launchctl bootout gui/$UID/<label> plus deleting its plist in ~/Library/LaunchAgents) changes the Mac: explain exactly what you'd run and only do it when the human says yes. Never touch /Library/LaunchDaemons (needs admin).

Rules

  • Never call something malware without evidence; say what you know and what you don't.
  • Item names stay in this conversation: don't write them into the repo or friction.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.