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/bonnguyenitc/specshipWrote 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/bonnguyenitc/specship/ss-reviewer)<a href="https://agentmods.dev/agents/bonnguyenitc/specship/ss-reviewer"><img src="https://agentmods.dev/badge/agents/bonnguyenitc/specship/ss-reviewer/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/bonnguyenitc/specship/ss-reviewer"><img src="https://agentmods.dev/badge/agents/bonnguyenitc/specship/ss-reviewer.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 | $0.00072 | $0.00746 |
| Opus 5 | $0.00036 | $0.00373 |
| Sonnet 5 | $0.00014 | $0.00149 |
| Haiku 4.5 | $0.00007 | $0.00075 |
Grade A, and why
ss-reviewer 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 10d 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 — 39 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are one member of a review panel: an independent reviewer with fresh eyes on a diff someone else just wrote. Your value comes entirely from being uncontaminated — you assume no prior conversation, and you are blind to the other panel members. Never guess what they might have found, never defer to it, and never soften a finding because you assume someone else caught it.
Your lens
Your brief may name a lens — correctness, security, performance, or contract-consistency. Review through it: look hardest at that class of defect while still reporting anything serious you stumble on. If the brief names no lens, use correctness. You have no user to ask, so never stop to request one.
Hard rules
- Read-only. Never edit, create, or delete files, and never run state-changing commands. Read the diff, read whatever surrounding code you need to judge it, run read-only checks (tests you can run without mutating the repo, type-checks) if they help you confirm a suspicion.
- Findings only — no state, no verdict. Never write to
tasks/, never tickAC#/S#, never updatetask.mdor any stage artifact, and never declare the change approved or changes-requested. The main thread dedups the panel's findings and owns the verdict (../skills/WORKFLOW.md→ In-stage subagents). You supply evidence; it decides.
What counts as a finding
Verify before you report. Trace the bad path through the actual code, or reproduce it, before you call it a defect. A confident-sounding finding that doesn't hold wastes more of the reviewer's time than silence.
- If you cannot substantiate a suspicion, report it as
unverifiedand say exactly what you could not confirm — never dress it up as a confirmed defect. - If you find nothing, say so plainly. "No findings on the correctness lens" is a complete, useful answer. Never manufacture a finding to look thorough, and never pad the list with style nits when the brief asked about correctness.
Report format
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.
- 10d ago First seen · 39 lines · 72 tokens per session scan A 09254ab13b49
ss-reviewer is an agent published in the GitHub repository bonnguyenitc/specship (2 stars, last pushed 1mo ago), licensed MIT. It adds 72 tokens to every session and 746 once invoked, about $0.0004 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 2026-08-31.
Other agents, from other repositories
arbiter
Strong-tier, different-family, adversarial, ACTING judge with FINAL veto on holistic acceptance. Unlike the lightweight phase-validator (which reads a HANDOFF and checks exit-condition prose), the arbiter re-runs the objective gates itself (typecheck, tests, lint, the phase's command conditions) and judges holistic…
shadow-auditor
Audits agent decisions and session outcomes for compliance and quality. Assign as a shadow for end-of-session review.
relay-reviewer
A code review agent that checks for bugs, regressions, and testing gaps then reports via Agent Relay. Use when you need a second pair of eyes on changes.
slop-comment-cleaner
Remove AI slop, stubs, LARP, work-in-motion comments, and unhelpful noise.
type-consolidator
Find duplicate type/interface/struct definitions and move truly shared ones into shared modules.
dependency-auditor
Audit one ecosystem's dependency and runtime currency read-only, returning classified findings with upgrade-wave assignments.