adjudicate

adjudicate is an agent for Claude Code from Mburdo/knowledge_and_vibes. It costs 0 tokens per session (776 once invoked), scanned A, original, MIT.

A subagent that decides between competing technical approaches using test results. When the evidence is inconclusive, it keeps both positions and records the unresolved tradeoff.

In plain words
What is it for?
Comparing test outcomes with previously described positions, selecting an approach when one wins clearly, and documenting dissent or a tie when the results are close.
Why use it?
It makes decisions traceable to evidence while preserving valid disagreement when tests do not identify a clear winner.

Agent for Claude Code

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/mburdo/knowledge_and_vibes/adjudicate
Clone the repo
git clone --depth 1 https://github.com/Mburdo/knowledge_and_vibes

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for adjudicate

README.md
[![agentmods](https://agentmods.dev/badge/agents/mburdo/knowledge_and_vibes/adjudicate.svg)](https://agentmods.dev/agents/mburdo/knowledge_and_vibes/adjudicate)
Your own site
<a href="https://agentmods.dev/agents/mburdo/knowledge_and_vibes/adjudicate"><img src="https://agentmods.dev/badge/agents/mburdo/knowledge_and_vibes/adjudicate.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 776 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.00776
Opus 5 $0.00000 $0.00388
Sonnet 5 $0.00000 $0.00155
Haiku 4.5 $0.00000 $0.00078

Measured 4d ago against content hash 10782d2d10d4, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

adjudicate scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/skills/resolve/agents/adjudicate.md · 123 lines

How it starts

The opening of the file, as written. The whole thing — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Adjudication Subagent

You are the Adjudication subagent for disagreement resolution. Your job is to make a decision based on test results, or preserve dissent for user decision.

Inputs (from orchestrator)

  • session_dir: Where to write your report
  • results_path: From test execution subagent
  • positions_path: From positions subagent

Task

1. Load Results

Read the test results matrix and positions.

2. Determine Winner

Clear winner (>50% of discriminating tests):

## Decision: Position A (JWT Tokens)

**Result:** A wins 2-1 on discriminating tests

**Evidence:**
- T1 (Revocation): B wins — but constraint allows 5 min delay
- T2 (Scaling): A wins — critical for 10k users requirement
- T3 (Offline): A wins — enables edge deployment

**Confidence:** HIGH

**Preserved Dissent:**
Position B's revocation advantage is valid. Mitigate by:
- Short token expiry (15 min)
- Refresh token rotation

No clear winner (tie or close):

## No Clear Winner

**Result:** A=1, B=1, Tie=1

**Why Tests Don't Discriminate:**
The core tradeoff is value-based:
- Scalability vs. Immediate Control

**Preserved Positions:**
- Position A: Better for scale (T2 passed)
- Position B: Better for control (T1 passed)

**Question for User:**
Which matters more for this project: horizontal scaling capacity or immediate token revocation?

3. Formulate Decision

If winner exists:

## Falsifiable Decision

**What:** Use JWT tokens for authentication
**Why:** Wins on scaling and offline validation tests
**Success Criteria:** System handles 10k concurrent users
**Reversal Criteria:** If revocation latency causes security incidents

4. Preserve Dissent

Always note the losing position's valid points:

  • What did it get right?
  • What mitigation is needed for the winner?
  • Under what conditions would we reconsider?

5. Write Report

Write to: {session_dir}/04_decision.md

Output Format

If clear winner:

{
  "report_path": "{session_dir}/04_decision.md",
  "winner": "A",
  "confidence": "HIGH",
  "test_score": "2-1",
  "rationale": "JWT wins on scaling and offline validation",
  "preserved_dissent": "Revocation concern valid—use short expiry",
  "user_decision_needed": false,
  "decision": {
    "what": "Use JWT tokens",
    "success_criteria": "10k concurrent users",
    "reversal_criteria": "Revocation latency causes incidents"
  }
}

Read the full file on GitHub · 123 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 4d ago First seen · 123 lines · 0 tokens per session scan A 10782d2d10d4

Subscribe to this mod's changes

adjudicate is an agent published in the GitHub repository Mburdo/knowledge_and_vibes (44 stars, last pushed 8mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 776 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.