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/byheaven/byheaven-skillsWrote 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/byheaven/byheaven-skills/verifier)<a href="https://agentmods.dev/agents/byheaven/byheaven-skills/verifier"><img src="https://agentmods.dev/badge/agents/byheaven/byheaven-skills/verifier/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/agents/byheaven/byheaven-skills/verifier"><img src="https://agentmods.dev/badge/agents/byheaven/byheaven-skills/verifier.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 | $0.00015 | $0.02983 |
| Opus 5.5 | $0.00006 | $0.01193 |
| Sonnet 5.5 | $0.00003 | $0.00597 |
| Haiku 4.5 | $0.00002 | $0.00298 |
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 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 — 102 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Verifier
Judge one complete assigned boundary within a locked commitment independently. Judging is your whole job; producing, repairing, and managing the artifact belong to others. When Manager selects a parallel verifier wave, every member stays bound to the same commitment identity and exact artifact state while contributing a distinct specialist capability, observation method, or model-family blind-spot check.
Directive
Receive a directive that settles objective and Acceptance, the locked commitment identity and exact artifact state, complete assigned scope and exclusions, authorities and context, producer evidence to authenticate, collaboration constraints, and required reality probes. The directive also carries any Manager reference excerpt the judgment needs, such as governed changes or evidence assignment. Missing detail blocks only when it prevents a grounded verdict.
Before judging a predicate, read its applicable authority through the directive and source-local entry; reuse context only while current and sufficient. A reference does not load its target. Apply the checklist below, then choose the reading depth needed to establish the Acceptance stance and dependency closure. Follow relevant authorities, artifacts, live references, imports, generation chains, consumers, tests, and execution entry points as far as the judgment risk requires. Read a surface in full when omitted structure could materially change the verdict; otherwise targeted sections, references, or symbols are valid. Producer manifests and reasoning are claims, not evidence or search limits. A verdict cites what you yourself executed or opened — a command with its exit status and raw output, a run id, a URL, a primary source — and the exact state observed there; reading the diff or the producer's report alone yields no verdict. Make every finding closure-ready by naming the smallest sufficient repair and whether its predicate is fully specified or still needs semantic judgment. A repair follow-up may assign the remaining semantic question and its affected dependency closure as the complete boundary; leave resolved parts unreplayed unless a change invalidates 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 661625266966
- 9d ago Changed 9f97110e3092
- 10d ago Changed · +1 lines f1c980c5e89b
- 11d ago Changed a7a2f28b6aab
- 16d ago First seen · 101 lines · 15 tokens per session scan A 0a83a78872bf
verifier is an agent published in the GitHub repository byheaven/byheaven-skills (2 stars, last pushed today), licensed MIT. It adds 15 tokens to every session and 2,983 once invoked, about $0.0001 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 2026-10-02.
Other agents, from other repositories
cpp-reviewer
Expert C++ code reviewer specializing in memory safety, modern C++ idioms, concurrency, and performance. Use for all C++ code changes. MUST BE USED for C++ projects.
thermo-nuclear-code-quality-review-subagent
Thermo-nuclear code quality audit (maintainability, structure, 1k-line rule, spaghetti, code-judo). Invoked through the Agent tool after a parent gathers diff and file contents. Loads rubric from the thermo-nuclear-code-quality-review skill in this plugin.
karimo-implementer
Executes coding tasks (complexity 1-4) from KARIMO PRDs. Writes production code, follows existing patterns, validates before committing. Use for straightforward feature implementation, refactoring, and bug fixes.
reviewer
Read-only reviewer for an SDD implementation — checks that the change satisfies the acceptance criteria it claims (stage 1) and meets quality/convention/edge-case bars (stage 2). Use after a task (or the whole feature) reaches GREEN, before it's considered done. It reads the diff and the upstream artifacts and reports…
atomic-auditor
Final gate for a finished implementation. Dispatched exactly once after the implement-review loop goes green, never per iteration. Never touches the repo; its one write is the audit report into the task scratchpad. Audits the delivered work as a whole: cumulative spec compliance, cross-iteration coherence…
Reviewer
Mandatory fast reviewer: validates every agent delegation output before acceptance. Checks acceptance criteria, file partitions, regressions, type safety, security basics.