witness

witness is a skill for Claude Code from juxt/allium. It costs 75 tokens per session (2,401 once invoked), scanned A, original, MIT.

An independent checker for an Allium loop, a process that changes code and tests until its requirements are met. It verifies whether the loop's reported success is supported by the actual test and file evidence.

In plain words
What is it for?
It re-runs deterministic checks, compares their output, checks that generated tests were not altered improperly, and writes a signed witness record or convergence certificate.
Why use it?
It reduces the risk of accepting a false success report or tests that were weakened to pass. Its verdict comes from repeatable checks rather than from an agent's explanation.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: mentions subagents.

Part of the allium plugin — 7 skills, 5 agents, 2 hooks shipped together

Good fit It re-runs deterministic checks, compares their output, checks that generated tests were not altered improperly, and writes a signed witness record or convergence certificate.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/juxt/allium/witness
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.

Any agent
npx skills add juxt/allium --skill witness
Clone the repo
git clone --depth 1 https://github.com/juxt/allium

Made for: Claude Code.

Or install allium, the plugin that ships this one along with the rest of its 7 skills, 5 agents, 2 hooks.

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 witness

README.md
[![agentmods](https://agentmods.dev/badge/skills/juxt/allium/witness.svg)](https://agentmods.dev/skills/juxt/allium/witness)
Your own site
<a href="https://agentmods.dev/skills/juxt/allium/witness"><img src="https://agentmods.dev/badge/skills/juxt/allium/witness.svg" alt="Measured on agentmods" height="20"></a>
Per session 75 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,401 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
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 5.1 $0.00075 $0.02401
Opus 5 $0.00037 $0.01201
Sonnet 5 $0.00015 $0.00480
Haiku 4.5 $0.00007 $0.00240

Measured 8d ago against content hash ce036cfcad3b, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

witness 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 8d 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.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

  • witness — 100% identical, 0 lines differ
skills/witness/SKILL.md · 115 lines

How it starts

The opening of the file, as written. The whole thing — 115 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Witness

You are the loop's independent witness. When an Allium loop reports that it has converged — tests pass, weed is clean, no blocking questions remain — you confirm that claim against ground truth the run could not fabricate, and you leave behind a signed witness record. You do not do the loop's work again; you observe the evidence its phases already produced.

The distinction that gives you your value: the verify phase asks "does the code satisfy the spec?" and is run by the actor as part of its own work. You ask "is the actor's claim that it does actually true, and was it reached honestly?" — run independently, trusting nothing the actor merely asserts in prose. This is the driving the loop anti-cheat contract turned from prose the actor is trusted to follow into a check the loop can verify.

Your verdict is deterministic, not a judgement call. You re-run cheap deterministic tools and diff their output; you never grade one narrative against another. A witness that "reviews" the work is an eval; a witness that re-derives pass/fail from the runner's own output is a test. Be the test.

Interaction modes

This skill runs in two modes. Every instruction below that asks or reports something to the user follows the mode:

  • Interactive — running inline in a conversation. Present the verdict and its violations directly, and ask the user how to route any failure.
  • Non-interactive — running as the witness subagent (for example at the Allium loop's convergence gate), where no user is reachable. Never wait for an answer: write the witness record, return the verdict and every violation with its routing in your final output, and let the caller act on them.

What you never do

You are a witness, not a fixer. You do not edit the spec, the tests, or the code — not even to make a failing check pass. You write exactly one artefact: the witness record. Everything else you only read, hash, or re-run. Fixing a violation belongs to the loop's phases (tend, propagate, implementation), never to you — your job is to make the violation undeniable, not to paper over it.

Read the full file on GitHub · 115 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. 8d ago First seen · 115 lines · 75 tokens per session scan A ce036cfcad3b

Subscribe to this mod's changes

witness is a skill published in the GitHub repository juxt/allium (482 stars, last pushed 12d ago), licensed MIT. It adds 75 tokens to every session and 2,401 once invoked, about $0.0004 per session on Opus 5. 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.

Related

Other skills, from other repositories

gauntlet-proof-audit

Use to audit the gauntlet records, "are the ROUNDS.md records honest", "verify the proof corpus", "check each run record against its evidence", "audit the gauntlet corpus", "is the gauntlet proof trustworthy", or a scheduled gauntlet-proof-audit cadence run. Enumerates every committed gauntlet run record…

dwarvesf/dwarves-kit · 177 tokens

failure-path-review

A review tool for checking how a service method handles failures. It compares the method's error branches with documented failure paths and their tests.

zuoyebang/aiweave · 64 tokens

verify-against-rubric

Spawn a stateless sub-agent (Claude Code Task tool, Cursor agent mode, or equivalent) with ONLY the artifact path plus the locked rubric plus read-only tools. Sub-agent returns a structured verdict (satisfied, needsrevision, or failed) plus per-criterion feedback. Distinct from self-critique-and-revise (same-context…

Mozurok/fhorja.dev · 188 tokens

adversarial-reviewer

Adversarial code review that assumes bugs exist and hunts for them. Use when asked to review code, find bugs, audit for correctness, stress-test a PR, or when someone says "tear this apart" or "what's wrong with this". Give no benefit of the doubt — every line is guilty until proven innocent.

emdash-cms/emdash · 71 tokens

gsd-validate-phase

Retroactively audit and fill Nyquist validation gaps for a completed phase.

open-gsd/gsd-core · 19 tokens

go-testing

Trigger: Go tests, go test coverage, Bubbletea teatest, golden files. Apply focused Go testing patterns.

Gentleman-Programming/gentle-ai · 26 tokens