redteam
Skill31 starsChanged 32 days agoMITView on GitHub
Runs one adversarial review of a claim, plan, or answer. Use only for /redteam, "push back on this", "red-team this", or "stress-test this". Identifies failure mechanisms and the cheapest way to test them. For broader ideation use fanout; for source verification use prove.
12 other projects use a version of this.
Adopted by
Most recently taken up by
- agentscope-ai/OpenJudgetheir own version, first seen Oct 25, 2026
- rivendale/ai-redteamtheir own version, first seen Oct 17, 2026
- captkernel/claude-thinking-toolstheir own version, first seen Oct 15, 2026
- vansvost-arch/claude-skillstheir own version, first seen Oct 15, 2026
- pr-purgatory/Skillstheir own version, first seen Oct 12, 2026
- modelscope/OpenJudgetheir own version, first seen Oct 9, 2026
Used with
Projects with it also use premortem.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
neurofooSkillpremortem
Pre-mortem analysis that imagines a plan has failed, then works backward to identify causes and preventions. Use before launches, major decisions, or risky initiatives to surface hidden risks.
Also in 3 projects that use this
Other versions12 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 12
- modelscope/OpenJudge★ 1.1ksame name
- agentscope-ai/OpenJudge★ 1.1ksame name
- neurofoo/agent-skills★ 111same name
- ktol1/RedTeam-Agent★ 67same name
- Hassaan146/claude-skills★ 18same name
- CarbeneAI/Forge★ 9same name
- pr-purgatory/redteam-skill★ 2same name
- rivendale/ai-redteam★ 2same name
- captkernel/claude-thinking-tools★ 0same name
- mj-deving/pai-skills★ 0same name
2 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
We split this skill into the rules it gives an agent
4 rules. 0 are common, 4 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 9 edited versions we read keep these.
4
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
- Distinctive: line 9
Read a referenced artifact and its project's current state and instructions to EOF.
- Distinctive: line 13
Alternative explanation: if supported; omit rather than invent a decoy.
- Distinctive: line 18
If the target is a file this session wrote and the fix is within the authorized scope, apply it and report the change.
- Distinctive: line 20
Run one pass. Be direct, distinguish evidence from hypotheses, and say when the target survives.
“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.
Files
Red-team review
- Identify the target from the supplied passage, file, or previous answer. Ask once if ambiguous.
- Read a referenced artifact and its project's current state and instructions to EOF. Include existing decisions and constraints in the review.
Attempt to disprove the target. Report:
- Strongest objection: the most consequential plausible failure, with its mechanism and evidence.
- Critical assumption: what must hold for the target to work.
- Alternative explanation: if supported; omit rather than invent a decoy.
- Cheapest disproof: a concrete check that could settle the objection.
- Verdict: survives, needs revision, or abandon, with the most important fix.
Opinion mode, only when the user gives their own opinion together with material ("here's what I think, here's the data"): do step 2 even if no file is named, so current project state outranks a stale excerpt. Replace the first three report items with the three strongest arguments against the opinion, each quoting the line of the material or project files that supports it, and use nothing from general knowledge; if the material supports fewer than three, give fewer and say so. Cheapest disproof and verdict stay. Any other target gets the ordinary report.
- If the target is a file this session wrote and the fix is within the authorized scope, apply it and report the change. Otherwise deliver the review without editing someone else's artifact.
Run one pass. Be direct, distinguish evidence from hypotheses, and say when the target survives. Do not manufacture flaws or claim an unrun check proves the verdict.
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.