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.
git clone --depth 1 https://github.com/MadAppGang/magusWrote 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.
[](https://agentmods.dev/commands/madappgang/magus/qa)<a href="https://agentmods.dev/commands/madappgang/magus/qa"><img src="https://agentmods.dev/badge/commands/madappgang/magus/qa/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/madappgang/magus/qa"><img src="https://agentmods.dev/badge/commands/madappgang/magus/qa.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00051 | $0.02029 |
| Opus 5.5 | $0.00020 | $0.00812 |
| Sonnet 5.5 | $0.00010 | $0.00406 |
| Haiku 4.5 | $0.00005 | $0.00203 |
Grade A, and why
qa 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 today.
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.
How it starts
The opening of the file, as written. The whole thing — 147 lines — stays where its author put it; the contents beside it link to each section on GitHub.
If --spec or --contract is missing, or a path does not exist, stop with
BLOCKED — <what is missing>. Never search the repo for a spec to substitute.
1. Check for child mode first, as "Inside a claudish session (child mode)" in that
skill describes. `list_models` still answers in child mode, so it cannot tell you.
In child mode you are already the delegate: go to step 4, and the report says
`writer: internal (nested claudish session; blindness by agent contract)`.
Otherwise, `list_models()`: if the tool is unavailable or errors, the writer is
internal (go to step 4) and the report says `writer: internal (claudish unreachable)`.
2. `search_models({ query: "<tier word>" })` with the `--model` value (default `gpt`).
Pick the top-tier model of that family from the live results. The tier the user
named is a hard constraint: if the catalog has no model matching it, stop and show
the live alternatives from the search; never pick a lower tier because its name
is closer. Record the chosen id.
3. External writer:
```
create_session(
model = <resolved id>,
agent = "dev:qa-engineer",
work_dir = "${SESSION_PATH}/qa/sandbox",
timeout_seconds = 900,
prompt = <the writer prompt below, with sandbox-relative paths>
)
```
`work_dir` is the sandbox, so the child session's tools cannot reach the
implementation: blindness is a property of the file system, not of the prompt.
Then wait for the channel `completed` event and call `get_output(session_id)`;
on `failed` or `timeout` call `get_diagnostics(session_id)`, record it, and fall
through to step 4 with `writer: internal (external failed: <reason>)`.
4. Internal writer:
```
Agent(
subagent_type: "dev:qa-engineer",
run_in_background: false,
description: "blind tests: <feature>",
prompt: <the same writer prompt, sandbox paths>
)
```
The subagent runs in its own context window; the prompt names only sandbox
paths, and the agent's own contract refuses Reads outside them.
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.
- today Changed · +4 lines 30b7fd6f8da4
- 22d ago Changed · -1 lines 70caabe0c067
- 1mo ago First seen · 144 lines · 51 tokens per session scan A 2b626ae8a919
qa is a command published in the GitHub repository MadAppGang/magus (13 stars, last pushed today), licensed MIT. It adds 51 tokens to every session and 2,029 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-09-15.
Other commands, from other repositories
prototype
You are building a proof-of-concept for the current Grainulator sprint. Read CLAUDE.md for sprint context and claims.json for existing research claims.
qa-changes
This skill should be used when the user asks to "QA a pull request", "test PR changes", "verify a PR works", "functionally test changes", or when an automated workflow triggers QA validation of code changes. Provides a structured methodology for setting up the environment, exercising changed behavior, and reporting…
verify
Run repository verification using the verification-loop skill.
modernize-transform
Transform one legacy module to the target stack — idiomatic rewrite with behavior-equivalence tests.
test-coverage
Analyze test coverage and identify the highest-value gaps to fill.
flowspec.validate
../../templates/commands/flow/validate.md.