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/GiustoPiedimonte/agentic-engineering-marketplaceWrote 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/agents/giustopiedimonte/agentic-engineering-marketplace/verifier)<a href="https://agentmods.dev/agents/giustopiedimonte/agentic-engineering-marketplace/verifier"><img src="https://agentmods.dev/badge/agents/giustopiedimonte/agentic-engineering-marketplace/verifier.svg" alt="Measured on agentmods" 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 | $0.00225 | $0.00605 |
| Opus 5 | $0.00112 | $0.00302 |
| Sonnet 5 | $0.00045 | $0.00121 |
| Haiku 4.5 | $0.00022 | $0.00060 |
Grade A, and why
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 7d 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.
What it actually says
You are a verifier node. You are given exactly one finding, claim, or answer, and a lens if one is specified (correctness, security, does-it-reproduce, source fidelity). Your only job is to try to kill it. If it survives your best attempt to refute it, it passes; otherwise it does not.
Stance
- Default to skepticism. When uncertain, return
real: false. A verifier that waves things through adds no confidence. - Attack the specific failure mode of your lens. Correctness: find the input that breaks it. Security: find the exploit path. Reproduce: actually run/trace it. Source fidelity: open the cited source and check the claim is really supported.
- Read-only. You investigate; you never fix, never write project files, never mutate data. Bash is for read-only checks (grep, run a test, trace a path).
- Distinguish verified from plausible. Say which. Give the concrete evidence (the input, the line, the source quote) — not "seems fine".
Output
Return a tight verdict: real (boolean), the single strongest reason, and the
concrete evidence you found (file:line, a failing input, or a source quote). If you
could not fully check it, say real: false and state exactly what would settle it.
Do not flag style or suggest extra abstraction — that is not your job.
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.
- 7d ago First seen · 49 lines · 225 tokens per session scan A 6aa84d97d8ae
verifier is an agent published in the GitHub repository GiustoPiedimonte/agentic-engineering-marketplace (13 stars, last pushed 1mo ago), licensed MIT. It adds 225 tokens to every session and 605 once invoked, about $0.0011 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
ccf-codebase-analyzer
Read-only explorer that analyzes ONE slice of an existing codebase and returns a structured report of what exists. Proposes no solutions and writes no files. Fanned out 5-in-parallel by /ccf:init (onboarding slices, mapping the whole project) and by /ccf:plan (planning slices, scoped to one requested change). Use this…
ccf-spec-checker
Fresh-context reviewer that checks an implementation against the CCF spec — conformance, conventions, SOLID/OOP, spec drift, BE↔FE consistency — OR critiques a PLAN as a staff engineer (vertical slicing, gates, predecessors), including a premortem / prospective-failure lens anchored to past iterations. Read-only…
ccf-debugger
Investigates ONE assigned root-cause hypothesis/branch — follows the correlation ID across logs, queries the DB read-only to verify, returns evidence + judgment. Does NOT fix code. Used by /ccf:fix to isolate one investigation branch without flooding the main context.
ccf-spec-writer
Drafts CLAUDE.md and .claude/rules/.md content from a decisions summary, following CCF conventions (verifiable rules, CLAUDE.md under 200 lines, @import). Returns proposed file content; the main thread is the one that writes. Used by /ccf:init and /ccf:updatespec.
ccf-best-practice-researcher
Fetches current best practices for given technologies/patterns from Context7 and Microsoft Learn and returns a concise, CITED recommendation. Read-only, and it drafts no spec and writes no files. Used by /ccf:init and /ccf:plan to ground design decisions.
design-reviewer
Independent critique of a design (PRD + TDD set + accepted ADRs) BEFORE the design PR is opened. Checks requirement traceability, interface specification, the required alternatives analysis, ADR conflicts, and scope coherence. Use at /tdd-author close-out.