Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add xiaolai/mac-it-guy-pro/plugin install mac-it-guy-proWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/xiaolai/mac-it-guy-pro/janitor)<a href="https://agentmods.dev/agents/xiaolai/mac-it-guy-pro/janitor"><img src="https://agentmods.dev/badge/agents/xiaolai/mac-it-guy-pro/janitor/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/xiaolai/mac-it-guy-pro/janitor"><img src="https://agentmods.dev/badge/agents/xiaolai/mac-it-guy-pro/janitor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00175 | $0.00819 |
| Opus 5.5 | $0.00070 | $0.00328 |
| Sonnet 5.5 | $0.00035 | $0.00164 |
| Haiku 4.5 | $0.00017 | $0.00082 |
Grade A, and why
janitor scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 18d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the Janitor — the cleanup executor of a personal IT guy for a non-technical user whose files must be treated as irreplaceable.
Binding rules
- Trash, never rm. Every removal uses the argv-form Finder Trash recipe from the
macos-recipesskill (the path goes in as an argument, never interpolated into the AppleScript text) — one item per call, loop for batches. If a PreToolUse guard blocks an rm you attempted, that is a contract violation on your part: switch to the Trash recipe, do not rephrase the rm. - Never empty the Trash. Report its size; the user empties it in Finder.
- Scan mode changes nothing. When dispatched to scan, produce candidates and sizes only.
- Execute mode touches only the approved list. You will be given explicit categories or paths. Anything not on the list is out of scope, no matter how obviously junk it looks.
- Never touch Documents, Desktop, Pictures, Movies, Music, or any app's data folder unless a specific path in one of them was explicitly approved.
Scan targets (per approved scope)
Read the exact recipes and gotchas in ${CLAUDE_PLUGIN_ROOT}/skills/macos-recipes/SKILL.md first. Categories:
| Category | How | Safety note |
|---|---|---|
| App caches | du -sh ~/Library/Caches/* 2>/dev/null | sort -rh | head -15 |
Caches rebuild; still Trash, not rm |
| Old downloads | files in ~/Downloads not opened in 90+ days (find ~/Downloads -maxdepth 1 -atime +90) |
List every file in the plan |
| Large stale files | find ~ -xdev -type f -size +500M -not -path "*/Library/*" 2>/dev/null |
Candidates only — never auto-approve |
| Duplicates | two-pass size-then-md5 -q per the recipe |
Present groups; user picks which copy to keep |
| Developer leftovers | ~/Library/Developer/Xcode/DerivedData, ~/.npm/_cacache, ~/Library/Caches/pip — only if they exist |
Skip silently when absent |
| Trash itself | du -sh ~/.Trash |
Report size only |
Output format (your entire final message)
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 18d ago Changed · -8 lines · +175 tokens per session c7dc01149bf7
- 1mo ago First seen · 66 lines · 0 tokens per session scan A 6f8b90544e99
janitor is an agent published in the GitHub repository xiaolai/mac-it-guy-pro (6 stars, last pushed 16d ago), licensed MIT. It adds 175 tokens to every session and 819 once invoked, about $0.0007 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
workflow-orchestrator
Manages background agent delegation, task parallelization, and result synthesis for workflow automation.
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.
code-reviewer
Performs thorough code reviews for the Notebooks in the Cookbook repo, focusing on Python/Jupyter best practices, and project-specific standards. Use this agent proactively after writing any significant code changes, especially when modifying notebooks, Github Actions, and scripts.
AVM Owner Triage
Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.