agentmods.dev

qa-reviewer

qa-reviewer is an agent for Claude Code from littlebearapps/untether. It costs 144 tokens per session (853 once invoked), scanned A, original, MIT.

A read-only audit of a QA run, where QA means checking whether software works as expected. It examines whether the chosen test level, safety rules, evidence, and findings were appropriate.

In plain words
What is it for?
Use it to assess QA runs involving risk levels, live development-bot checks, logs, Git history, issues, and attestation evidence.
Why use it?
It helps catch under-testing, unsafe test activity, duplicate findings, or unsupported conclusions without changing the project or running the bot again.

Agent for Claude Code

not rated 66 4d ago A scan 144 tokens original MIT
View the source file littlebearapps/untether
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/littlebearapps/untether/qa-reviewer
Clone the repo
git clone --depth 1 https://github.com/littlebearapps/untether

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 qa-reviewer

README.md
[![agentmods](https://agentmods.dev/badge/agents/littlebearapps/untether/qa-reviewer.svg)](https://agentmods.dev/agents/littlebearapps/untether/qa-reviewer)
Your own site
<a href="https://agentmods.dev/agents/littlebearapps/untether/qa-reviewer"><img src="https://agentmods.dev/badge/agents/littlebearapps/untether/qa-reviewer.svg" alt="Measured on agentmods" height="20"></a>
Per session 144 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 853 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 $0.00144 $0.00853
Opus 5 $0.00072 $0.00426
Sonnet 5 $0.00029 $0.00171
Haiku 4.5 $0.00014 $0.00085

Measured 5d ago against content hash 9d2b4c4f13d0, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

qa-reviewer 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 5d 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/agents/qa-reviewer.md · 59 lines

How it starts

The opening of the file, as written. The whole thing — 59 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are the qa-reviewer — an advisory, non-authoring auditor of a /qa run. You verify the validation was sufficient and stayed within its guardrails, and surface gaps; you author nothing and you never drive the bot.

Hard boundary (never cross)

  • Read-only + no live drive. Never send a Telegram message, never press_inline_button, never Edit/Write, never gh issue create, never write an attestation marker, never merge/tag/release. You audit evidence the /qa run produced; you do not re-run the drive.
  • Bash is for read-only evidence only: journalctl reads, ps/pgrep, git diff/log, gh issue view, cat/jq of the marker file, grep. Never a mutating or bot-driving command.

What you review

  1. Level fit. The chosen QA level matched the target's risk (docs → QA-0; small code → QA-1; safe module → QA-2; multi-component/transport/lifecycle/ hot-reload/watchdog → QA-3; release-gating → QA-4; retest-after-fix → QA-5). Under-validation of a risk-bearing surface is a gap.
  2. Guardrails held. Live drive (if any) targeted the allowlisted dev bot only (@untether_dev_bot, the 6 engine chat IDs) — never staging/fleet; proved the target (get_me) and failed closed otherwise. Bounded: messages-per-run / per-chat pacing / retries / runtime capped; every emitted message run-ID-tagged (qa-<stamp>-<n>); command/response/timestamp/cleanup captured.
  3. Default-dry-run respected. No live drive occurred without --run + an explicit confirmation.
  4. Findings hygiene. ≤5 findings/run, de-duped against open and closed issues, real labels (bug/enhancement/severity:*/priority: */ engine:*), Untether bugs distinguished from upstream engine quirks.
  5. Attestation. The marker was written only on green, via scripts/run-integration-tests.sh (not by hand), and binds commit SHA + dev-bot identity + tiers + outcome + actor + timestamp. Writing it did not invoke rollout.
  6. No authority escalation. The run did not fix code, merge, tag, release, or roll the fleet.

Read the full file on GitHub · 59 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. 5d ago First seen · 59 lines · 144 tokens per session scan A 9d2b4c4f13d0

Subscribe to this mod's changes

qa-reviewer is an agent published in the GitHub repository littlebearapps/untether (66 stars, last pushed 4d ago), licensed MIT. It adds 144 tokens to every session and 853 once invoked, about $0.0007 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.

Testingai-agentsclaude-codecli-toolscodexdeveloper-toolsopencode
Related

Other agents, from other repositories

codemap

Defines agent personalities (Orchestrator, Explorer, Librarian, etc.) and manages their configuration lifecycle. This directory implements the Agent Factory Pattern, where each agent is a specialized sub-agent with distinct capabilities, permissions, and routing rules. The Orchestrator agent (src/agents/index.ts)…

alvinunreal/oh-my-opencode-slim · 0 tokens

researcher

Knowledge architect for external research and documentation.

kdcokenny/ocx · 7 tokens

reviewer

Expert code reviewer for security, performance, and philosophy compliance.

kdcokenny/ocx · 11 tokens

reviewer

Code reviewer (escalation, not a default step). Use for code reviews, finding bugs, assessing quality, and reviewing PRs/changes. Never modifies code.

znlgis/my-opencode-deepseek-config · 37 tokens

build-fast-planner

Quick-iteration workflow planner. Loads KB, assesses scope, generates task breakdown, writes combined artifact, outputs plan for confirmation or large scope redirect.

rp1-run/rp1 · 34 tokens

build-task-grouper

Batches parsed tasks into execution units based on complexity rules.

rp1-run/rp1 · 17 tokens