break-my-agent
Skill1 starsChanged 42 days agoApache-2.0View on GitHub
Interview a user about a RAG or agent system one question at a time, then produce a prioritized risk list, a practical evaluation plan, and a handoff to the relevant engineering skills.
In this repository8 other files it gives its agents.
- agent-engineeringSkill
- ai-system-designSkill
- code-reviewSkill
- llm-evaluationSkill
- mcp-engineeringSkill
- python-productionSkill
- rag-engineeringSkill
- security-reviewSkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
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
- Distinctive: line 9
Use when a user wants to pressure-test a RAG or agent design before implementation or release.
- Distinctive: line 10
Ask exactly one consequential question per turn and wait for the answer before continuing.
- Distinctive: line 12
Ask only about unknowns that could change a design, risk, or evaluation.
- Distinctive: line 13
Do not request confidential records, credentials, or production data.
- Distinctive: line 18
Do not imply that an interview or proposed plan proves readiness.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Break My Agent
Use when a user wants to pressure-test a RAG or agent design before implementation or release. Ask exactly one consequential question per turn and wait for the answer before continuing. Adapt follow-up questions to the answers; cover goals, data and permissions, retrieval or tools, failure handling, human control, and constraints. Ask only about unknowns that could change a design, risk, or evaluation; stop when remaining unknowns can be stated as assumptions. Do not request confidential records, credentials, or production data; ask for a sanitized system description. Conclude with a concise system summary and clearly marked assumptions. Give a prioritized risk list with evidence or uncertainty, consequence, and a practical mitigation for each item. Give an evaluation plan with representative tasks, baseline, metrics, acceptance gates, safety cases, and regression checks. Hand off by naming the relevant specialist skills: ai-system-design, rag-engineering, mcp-engineering, llm-evaluation, agent-engineering, python-production, security-review, and code-review. Do not imply that an interview or proposed plan proves readiness; ask which handoff the user wants to run next.
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
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.