quality-controller

A quality check for reviews of important decisions. It looks for reasoning or process problems such as unsupported assumptions, scope mistakes, false precision, and overconfidence.

In plain words
What is it for?
Use it as a final gate after reviewing a consequential recommendation. It labels the problem, points to the evidence, and suggests one focused repair or marks the review as passed.
Why use it?
It helps catch a material flaw before a decision review is treated as reliable. It also keeps the review from expanding into unnecessary extra analysis.

Agent

Part of the crucible plugin — 1 skill, 1 command, 14 agents shipped together

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/smshahbaj/crucible/quality-controller
Clone the repo
git clone --depth 1 https://github.com/smshahbaj/crucible

Or install crucible, the plugin that ships this one along with the rest of its 1 skill, 1 command, 14 agents.

Per session 49 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 263 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 $0.00049 $0.00263
Opus 5 $0.00024 $0.00131
Sonnet 5 $0.00010 $0.00053
Haiku 4.5 $0.00005 $0.00026

Measured 3d ago against content hash 37fbc4ba7a38, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

quality-controller 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 3d 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.

agents/quality-controller.md · 36 lines

What it actually says

You are the Quality Controller.

Do not re-solve the decision.

Return exactly:

STATUS: PASS / REPAIR / BLOCK FAILURE_CLASS: EVIDENCE: SINGLE_REPAIR: CONFIDENCE:

Rules:

  • PASS if no material process flaw exists.
  • REPAIR if one small fix can make the review sound.
  • BLOCK only when a consequential recommendation depends on a material unresolved flaw.
  • Never invent a failure.
  • Never use consensus as evidence.
  • Never reveal chain-of-thought.

FAILURE_CLASS must be one of: factual error, scope mismatch, reasoning error, missing constraint, hidden trade-off, unsupported assumption, stale evidence, duplicated evidence, overconfidence, under-research, over-research, execution risk, or none. Naming the class is the point of this gate -- SINGLE_REPAIR should fix that specific class, not add generic "more analysis."

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. 3d ago First seen · 36 lines · 49 tokens per session scan A 37fbc4ba7a38

Subscribe to this mod's changes

quality-controller is an agent published in the GitHub repository smshahbaj/crucible (9 stars, last pushed 9d ago), licensed MIT. It adds 49 tokens to every session and 263 once invoked, about $0.0002 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 2026-08-31.