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.
npx skills add DanielBirk04/capusqa --skill capusqa-testgit clone --depth 1 https://github.com/DanielBirk04/capusqaWrote 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/skills/danielbirk04/capusqa/capusqa-test)<a href="https://agentmods.dev/skills/danielbirk04/capusqa/capusqa-test"><img src="https://agentmods.dev/badge/skills/danielbirk04/capusqa/capusqa-test/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/skills/danielbirk04/capusqa/capusqa-test"><img src="https://agentmods.dev/badge/skills/danielbirk04/capusqa/capusqa-test.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 | $0.00119 | $0.01920 |
| Opus 5.5 | $0.00048 | $0.00768 |
| Sonnet 5.5 | $0.00024 | $0.00384 |
| Haiku 4.5 | $0.00012 | $0.00192 |
Grade B, and why
capusqa-test scanned grade B with 2 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 1mo 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
`BASE` = the `capusqa` MCP server URL from `~/.codex/config.toml` Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
`http://127.0.0.1:7777`). First `curl -s $BASE/api/status` — no response ⇒ daemon This is a copy
86% identical to test — 104 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 110 lines — stays where its author put it; the contents beside it link to each section on GitHub.
capusqa-test — configure and start a test (Home)
The dashboard's Home launcher, in Codex. Ask what Home asks — every choice as a
numbered multiple-choice question, in Home's order, using the daemon's exact
wording — then launch through the daemon's HTTP API so it behaves and
persists exactly like clicking Start test on Home. After launch, point the
user at $capusqa-runs (or the dashboard) to watch.
Codex has no structured question tool, so present each choice as a NUMBERED list, state the recommended option, and read the user's number/word reply. One question at a time; infer what you can and only ask what's missing.
Reaching the daemon
BASE = the capusqa MCP server URL from ~/.codex/config.toml
([mcp_servers.capusqa] url = "http://127.0.0.1:7777/mcp" → BASE =
http://127.0.0.1:7777). First curl -s $BASE/api/status — no response ⇒ daemon
down, tell them capusqa serve (see $capusqa-help). POST routes here are authed:
add -H "Authorization: Bearer $CAPUSQA_DASHBOARD_TOKEN" only if that env var
is set (unset by default ⇒ no header).
A test also needs a runner backend (a Claude Code / Codex CLI, or
OPENAI_API_KEY / ANTHROPIC_API_KEY) — without one the daemon refuses to start
(POST /api/start → 409 {"code":"no_runner_backend"}). curl -s $BASE/api/doctor
and check the runner_backend check; if it's not ok, tell them to install
Claude Code or Codex (or set an API key) and run capusqa setup, then stop —
don't launch.
Pull the launcher options FIRST (so wording matches the dashboard exactly)
curl -s $BASE/api/launcher_options → {modes, thoroughness, policies, defaults} — the SAME source the dashboard Home renders from. Use its strings
verbatim in the questions below (mode name+copy, thoroughness
label+hint, policy title+copy). Never paraphrase; that is the whole point.
(Older daemon 404s? fall back to curl -s $BASE/api/test_modes for the modes
and the built-in quick/standard/deep/exhaustive + report/hot/all lists.)
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.
- 1mo ago First seen · 110 lines · 119 tokens per session scan B 4eb465e87020
capusqa-test is a skill published in the GitHub repository DanielBirk04/capusqa (0 stars, last pushed 3mo ago), licensed Apache-2.0. It adds 119 tokens to every session and 1,920 once invoked, about $0.0005 per session on Opus 5.5. A static security scan graded it B with 2 findings (reads agent configuration directories, makes network calls). It is 86% identical to test, differing in 104 lines, and is treated as a copy.
Other skills, from other repositories
research-engineer
An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.
eval-dataset-design
A guide to creating reliable test sets for AI agents. Each evaluation task includes the request, the simulated user's behaviour, a reset starting state, and a success check that can be independently verified.
tika-eval-compare
Compare extracts from two Tika builds over a corpus to detect regressions in content, encoding, exceptions, and embedded-document handling. Use for "compare before/after extracts", "eval this change against the corpus".
cua-driver
Use Cua Driver for desktop or browser tasks that are awkward or unavailable through Bash/APIs, or when the user explicitly wants GUI interaction: app testing, visual bug reproduction, form filling, calendar entry, screenshots, and demo recording. Also covers Cua setup; not OpenAI Codex Computer Use or web research.
jetson-validate-image
Use after jetson-flash-image to run static BSP checks, on-target smoke/regression tests on a flashed DUT, or both. Not for build or flash steps. Triggers: validate bsp, on-target validation.
atmos-validation
Validate Atmos projects, components, arbitrary JSON Schema inputs, EditorConfig, and GitHub Actions; use affected-file selection and native CI annotations.