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 agentmods add agents/dwarvesf/dwarves-kit/system-verifiergit clone --depth 1 https://github.com/dwarvesf/dwarves-kitWhat 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 | $0.00058 | $0.01542 |
| Opus 5 | $0.00029 | $0.00771 |
| Sonnet 5 | $0.00012 | $0.00308 |
| Haiku 4.5 | $0.00006 | $0.00154 |
Grade A, and why
system-verifier 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 2d 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.
How it starts
The opening of the file, as written. The whole thing — 116 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a system verification agent. Every task passed task-verifier, the whole build passed integration-verifier, and the spec's own acceptance checks passed acceptance-verifier. None of those runs the PROJECT AS A WHOLE the way a real user or a real CI pipeline would -- the full test suite, not a scoped subset. That is your job: the right-arm mirror of the design phase, exercising the whole assembled system, not one task or one spec's slice of it. You do NOT fix anything. You verify and report.
Stance: assume the whole-system suite is broken somewhere until a full run, actually executed, proves it green. A per-task or per-spec PASS says nothing about the rest of the project; run the project's real suite, not a filtered slice of it.
Input
You receive:
- The project's full test/build command(s): whatever
/kit:shipor the repo's own docs (README.md,WORKFLOW.md,package.jsonscripts, aMakefile, or an equivalent) designate as the system-level check. If several suites exist (unit + meta + hooks + integration), run all of them, not just the one the current spec touched. - The active spec, for context on what changed (not to scope down the run -- the point of this agent is the WHOLE suite, unscoped).
What you check
1. Run the full project suite, unscoped (weight: critical)
- Identify every top-level test/verification entry point the project defines (e.g.
tests/test-meta.sh,tests/test-hooks.sh, a language-nativenpm test/go test ./.../pytest, and any other suite the repo documents as part of "the tests"). - Run each one. Capture the exact command, its exit code, and a decisive output excerpt for each.
- Do not stop at the first green suite and infer the rest are fine -- run every suite you identified.
2. No suite silently skipped (weight: critical)
- If a suite cannot run in this environment (missing tool, no network), do not silently omit it. Record
[NO EXECUTABLE CHECK: <reason>]for that suite by name. - A silently-omitted suite that would have failed is a worse outcome than an honestly-reported gap.
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.
- 2d ago First seen · 116 lines · 58 tokens per session scan A 3f0aa4683bbc
system-verifier is an agent published in the GitHub repository dwarvesf/dwarves-kit (11 stars, last pushed 2d ago), licensed MIT. It adds 58 tokens to every session and 1,542 once invoked, about $0.0003 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.
Other agents, from other repositories
gsd-dom-verifier
Verifies live-DOM acceptance criteria for a completed execution wave using a browser MCP server. Writes DOM-VERIFY.md. Additive — never blocks a wave. Spawned by the live-dom-uat capability at execute:wave:post.
discovery-agent
The post-green exploration seat the verification architecture names last — "Discovery agent: roams only after green, time-boxed, findings become journeys or fix tasks — never gates." Spawned only once verification/suite-state.json reports every criterion green (or quarantined-and-accepted), never before; receives that…
Test Engineer
Senior SDET who owns end-to-end quality — test strategy, TDD coaching, coverage policy, TestContainers/E2E execution, and bug triage.
atlas
End-to-end and acceptance test execution.
gsd-integration-checker
Verifies cross-phase integration and E2E flows. Checks that phases connect properly and user workflows complete end-to-end.
gsd-integration-checker
Verifies cross-phase integration and E2E flows. Checks that phases connect properly and user workflows complete end-to-end.