adjudication-protocol
Skill0 starsChanged 4 months agoMITView on GitHub
Use when a contested set needs rulings - court procedure (advocate/skeptic/risk-officer), evidence requirements, sequential fallback, and escalation rules.
In this repository4 other files it gives its agents.
- merge-disciplineSkill
- oracle-forgingSkill
- spec-attackSkill
- tricolor-reportingSkill
The file
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 19
Escalation: only items still genuinely ambiguous AFTER execution go to the human — presented with both sides' strongest executed evidence …
≈ “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
Adjudication Protocol
Disputes are settled by executed evidence, not by argument quality.
Court composition (agent team, >3 contested items): advocate argues each item's strongest case; skeptic EXECUTES every repro command attempting to break the claim — armchair doubt does not count; risk-officer assesses blast radius and second-order effects. One shared task per item.
Evidence discipline:
- A ruling exists when its evidence file exists:
runs/<run-id>/verdicts/task-<task-id>.jsonwith the raw exit codes the skeptic observed. The TaskCompleted gate bounces completion without it. - Verdicts are computed from exit codes by script (
adjudicate.mjs), never voted on. The court's job is to produce better evidence, not consensus. - Judges receive the distilled contested set only (constitution C6) — raw corpus dilutes judgment.
Volatility discipline: teams have no resume. Timebox the court; flush every ruling to disk the moment it lands; a crash loses only the in-flight item. Degrade to sequential main-session adjudication reading the same files whenever teams are unavailable — announce the degradation once, without alarm.
Escalation: only items still genuinely ambiguous AFTER execution go to the human — presented with both sides' strongest executed evidence and a concrete question, never "what do you think?".
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.