verify-swarm
Skill1 starsChanged 4 months agoMITView on GitHub
Verify a change/claim/test with multiple independent agents and return a consensus verdict — not one agent's guess. Self-improving: logs verdicts vs reality. Use when the user says "verify with agents", "verify using multiple agents", "make sure it works", "run all tests and verify". Agentic OS (Level 1).
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
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 30
Never call PASS on assertion alone — require evidence per agent.
≈ “Common” is an estimate: at least 1 in 5 of the 308,232 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
verify-swarm
Independent multi-agent verification → majority verdict. Kills plausible-but-wrong "it works" claims.
Loop (every time)
- Before — read
<vault>/agentic-os/runs/verify-swarm.md; applylesson:(false-pass patterns).
Method
- Define the claim precisely + the observable pass criteria (test output, HTTP code, rendered UI, file state).
- Spawn 3 independent agents (Task tool,
general-purposeorproduction-validator), each told to try to REFUTE the claim — default to FAIL if evidence is missing. - Each returns
{verdict: PASS|FAIL, evidence, reason}. - Consensus: PASS only if ≥2/3 PASS with concrete evidence. Any FAIL with real evidence → investigate, don't average away.
After
Append run: claim, 3 verdicts, consensus, PASS/FAIL, lesson: (esp. if consensus was later proven wrong).
Output
Verdict line + per-agent evidence table. If FAIL, the specific failing evidence + suggested fix.
Never
- Never call PASS on assertion alone — require evidence per agent. Never run the 3 with identical prompts when the claim can fail multiple ways — give each a distinct lens (does it build / does it run / does it regress).
In this repository6 other files it gives its agents.
- deploy-verifySkill
- episode-packageSkill
- localize-humanizeSkill
- seo-guardSkill
- skill-onboardSkill
- video-cost-estimateSkill
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.