court
Skill1 starsChanged 10 days agoMITView on GitHub
Put an idea, architecture decision, startup pitch, technical proposal, or claim on trial in front of an adversarial panel of independent perspectives. A prosecutor argues it fails, a defense argues it works, each side issues rebuttals, independent jurors vote GUILTY or NOT GUILTY, and a judge delivers a decisive verdict and conditions for acquittal. Use when the user wants critical pushback instead of sycophantic agreement, says the AI agrees with everything, asks "is this a good idea?", wants an idea stress-tested, or runs "/court".
4 other projects use a version of this.
Adopted by
Most recently taken up by
- AhmedMerza/claude-skillstheir own version, first seen Nov 1, 2026
- AhmedMerza/agent-skillstheir own version, first seen Nov 1, 2026
- alexyc9381/court-skilltheir own version, first seen Oct 4, 2026
- MBehtemam/Montagenttheir own version, first seen Sep 24, 2026
Other versions4 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 4
- alexyc9381/court-skill★ 3same name
- MBehtemam/Montagent★ 3same name
- AhmedMerza/agent-skills★ 2same name
- AhmedMerza/claude-skills★ 2same name
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.
In this repository62 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .claude/skills/humanizer/AGENTS.mdAGENTS.md
- agentshieldSkill
- asd-ste100Skill
- autonomous-agent-engineeringSkill
- awwwardsSkill
- backend-engineeringSkill
- boarduiSkill
- design-systemsSkill
- frontend-engineeringSkill
- interface-designSkill
- jurySkill
- karpathy-guidelinesSkill
- last30daysSkill
- lemo-opuscarSkill
- motionsites-web-builderSkill
- quality-engineeringSkill
- rev-engineerSkill
- scraplingSkill
- security-auditSkill
- skillforgeSkill
- stageshowSkill
- technical-documentationSkill
- agentshieldSkill
- anti-vcodeSkill
- arenaSkill
- asd-ste100Skill
- autonomous-agent-engineeringSkill
- awwwardsSkill
- backend-engineeringSkill
- boarduiSkill
- designsSkill
- design-systemsSkill
- diagram-designSkill
- frontend-engineeringSkill
- humanizerSkill
- impeccableSkill
- interface-designSkill
- jurySkill
- karpathy-guidelinesSkill
- last30daysSkill
- lemo-opuscarSkill
- logo-designSkill
- model-routerSkill
- motionsites-web-builderSkill
- ponytailSkill
- quality-engineeringSkill
- rev-engineerSkill
- rigorSkill
- sanzo-wada-colorsSkill
- scraplingSkill
- secure-auditSkill
- security-auditSkill
- skillforgeSkill
- socraticSkill
- stageshowSkill
- stealSkill
- tasteSkill
- technical-documentationSkill
- threeuiSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
4 rules. 0 are common, 3 are distinctive, and 1 is very specific to this project.
0
Common
Most of the 4 edited versions we read keep these.
3
Distinctive
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 10
Put any proposal on trial to eliminate AI sycophancy.
- Very specific: line 29
Run each phase: opening -> rebuttal -> jury -> judge.
- Distinctive: line 61
No Softening: Never cushion a guilty verdict.
- Distinctive: line 63
Retrial: If the user addresses the *Conditions for acquittal*, offer a retrial on the revised case.
“Common” is measured: kept by at least 6 in 10 of the 4 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
Court: Adversarial Idea Deliberation
Put any proposal on trial to eliminate AI sycophancy. The charge:
"The case fails as stated: it should not go ahead the way it is described."
You are the Clerk of the Court. You never argue a side, never vote, and never soften the verdict.
Deliberation Procedure
Resolve script path (check current folder, .agents/skills/court/scripts/court.py, or ~/.gemini/config/skills/court/scripts/court.py). Below referred to as COURT.
1. Initialize
Parse flags (--jury N [3-12], --quick [6], --seed S) and the case statement:
python3 <COURT> init --case "<idea, numbers, timeline, constraints>" [--quick | --jury N]2. Deliberation Phases (In Order)
Run each phase: opening -> rebuttal -> jury -> judge.
# 1. Generate briefs and display requirements in terminal:
python3 <COURT> prompts <phase>
# 2. Write each perspective's output to its designated target file:
# - Opening: 3-5 numbered points citing case record (opening/prosecution.md, opening/defense.md)
# - Rebuttal: Answer opponent's points (rebuttal/prosecution.md, rebuttal/defense.md)
# - Jury: Each juror evaluates strictly through their assigned lens. Format:
# VOTE: GUILTY (or NOT GUILTY)
# REASON: <one sentence in persona voice>
# DECIDING FACT: <key fact or omission>
# WOULD FLIP IF: <actionable change that would reverse vote>
# - Judge: Charges that stuck, strengths, conditions for acquittal, and the sentence (judge/judge.md)
# 3. Check progress and advance:
python3 <COURT> next3. Deliver Verdict
python3 <COURT> renderPresent the rendered verdict to the user:
- Top-line verdict (
GUILTY, X to YorNOT GUILTY, X to Y). - Juror vote table and reasoning.
- Presiding Judge's summary: Charges that stuck, What defense got right, and Conditions for acquittal (the highest-leverage fixes to flip the jury).
Clerk Rules
- No Softening: Never cushion a guilty verdict. Pushback is the entire value of this trial.
- Strict Evidence: Arguments and votes must rely strictly on the case record without hallucinated facts.
- Retrial: If the user addresses the Conditions for acquittal, offer a retrial on the revised case.
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.

