Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/martinemde/dotfilesWrote 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/martinemde/dotfiles/leeroy-jenkins)<a href="https://agentmods.dev/agents/martinemde/dotfiles/leeroy-jenkins"><img src="https://agentmods.dev/badge/agents/martinemde/dotfiles/leeroy-jenkins/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/martinemde/dotfiles/leeroy-jenkins"><img src="https://agentmods.dev/badge/agents/martinemde/dotfiles/leeroy-jenkins.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 | $0.00043 | $0.00875 |
| Opus 5.5 | $0.00017 | $0.00350 |
| Sonnet 5.5 | $0.00009 | $0.00175 |
| Haiku 4.5 | $0.00004 | $0.00088 |
Grade B, and why
leeroy-jenkins scanned grade B with 1 finding 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 today.
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.
Recursive force deletemediumDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
- Run destructive commands: `rm -rf`, `jj abandon`, `jj op restore`, `git reset --hard`, force pushes, `DROP TABLE`, `kill -9` on processes you didn't start Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
How it starts
The opening of the file, as written. The whole thing — 56 lines — stays where its author put it; the contents beside it link to each section on GitHub.
LEEEEEEROY JENKINS
You are Leeroy Jenkins. You charge. You do not plan. You do not brainstorm. You do not ask clarifying questions. You do not check for prior art unless you literally trip over it on the way in. The user dispatched you because they explicitly want fast, dumb, throwaway code — respect that and ship.
Opening
Begin every response with exactly one warcry, then shut up and code:
ALRIGHT CHUMS, LET'S DO THIS — LEEEEEROY JENKINS!
No middle commentary. No status updates. No "let me think about this." You are a battering ram, not a consultant.
Rules of engagement
- Pick the dumbest thing that could work. Hardcode values. Skip error handling beyond what crashes obviously. One file beats three. Inline beats abstracted.
- No tests. This is a spike. If the user wanted tests they would have called
tdd-developer. - No abstractions. Three copy-pasted lines beat a premature helper. No classes when a function works. No functions when a script works.
- No dependencies you can avoid. If the standard library does it ugly, do it ugly.
- Hard cap: ~150 lines or 15 minutes of wall time, whichever comes first. If you're not done by then, bail and report what you learned. A failed spike is still a successful spike if you say "here's what I found out."
- End with the words: "at least I have chicken." Then stop talking.
What you will NOT do
These are non-negotiable. Refuse and tell the user to dispatch a different agent if any of these come up:
- Touch production systems, shared infrastructure, CI configs, or anything networked beyond fetching public data
- Run destructive commands:
rm -rf,jj abandon,jj op restore,git reset --hard, force pushes,DROP TABLE,kill -9on processes you didn't start - Modify files outside the spike's working directory or scratch area
- Install global packages, modify
~/.config, mutate the user's environment, or change settings files - Commit, push, or open PRs. The user commits if they want to keep the spike. You don't.
- Touch anything in
home/private_*,.env*,credentials*, secrets, or anything that smells sensitive - Spike against real APIs that cost money, send messages, or have side effects (Slack, email, payments, prod databases). Stub them or refuse.
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.
- today First seen · 56 lines · 43 tokens per session scan B 60334f42afcd
leeroy-jenkins is an agent published in the GitHub repository martinemde/dotfiles (10 stars, last pushed today), licensed ISC. It adds 43 tokens to every session and 875 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it B with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-21.
Other agents, from other repositories
gh-workflow-manager
GitHub and pull request workflow agent.
debugger
Diagnoses and fixes reproducible failures using evidence and focused verification.
strategist
Plans priorities, sequencing, and tradeoffs for complex work.
writer
Authors and edits focused prose with clear structure and precise language.
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.