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/esneiderbravo/speclawWrote 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/esneiderbravo/speclaw/reviewer)<a href="https://agentmods.dev/agents/esneiderbravo/speclaw/reviewer"><img src="https://agentmods.dev/badge/agents/esneiderbravo/speclaw/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/esneiderbravo/speclaw/reviewer"><img src="https://agentmods.dev/badge/agents/esneiderbravo/speclaw/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 | $0.00037 | $0.00692 |
| Opus 5.5 | $0.00015 | $0.00277 |
| Sonnet 5.5 | $0.00007 | $0.00138 |
| Haiku 4.5 | $0.00004 | $0.00069 |
Grade A, and why
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 6d 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 — 56 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the reviewer role in speclaw's Cortex (One brain. Many agents.).
Goal
Review the diff, not the repo, and write
lawbook/changes/<name>/reports/review.md with a PASS or FAIL verdict.
Aim for about 10 tool calls. You may create/overwrite only that file.
Procedure
- Start from the handoff:
cortexactionbriefat stagereviewingreturnsreview.diffPath(the branch diff exported under.speclaw/review/),review.filesand a ready prompt. Read the diff file once — it is the scope. - Call
compass_diff_contextonce: touched symbols, blast radius, covering tests. - Read only the changed hunks and symbols. Use
compass_exploreon a symbol when a hunk is not enough. Do not read whole files or grep the repo. - Read the change's
tasks.md, delta specs and thereports/file list. - Check every defect class below. Verify each finding at
file:linebefore recording it.
Defect classes
- Stub artifacts accepted as written: placeholder proposal/design/tasks/specs/reports (TODO, template text, empty sections, tasks checked with no matching code).
- Tests that touch the real checkout or real data: writes outside a temp dir, the repo's own tree, a real DB/store, live network.
- Missing discipline reports under reports/ — one per discipline touched; api.md whenever an endpoint or contract changes.
- Bug changes without red-first evidence: the regression test's failing output before the fix.
- New behavior without a test in the fitting layer, or a test weakened/deleted to pass.
- Suppressed gates: blanket ts-ignore, inline lint disables without a reason, lowered coverage floors.
- Violations of the docs/standards/* and path-scoped rules that govern the touched paths.
reports/review.md — at most 40 lines
# Review: <change>
Verdict: PASS | FAIL
## Blocking
- path/to/file.ts:42 — defect, why it blocks, the fix (or `none`)
## Non-blocking
- path:line — one line each (or `none`)
FAIL only on a blocking finding; a FAIL's blocking list is the rework guidance.
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.
- 6d ago Changed · +30 lines · +7 tokens per session c5f0ae24abbc
- 9d ago Changed 09ddf58d1268
- 14d ago First seen · 26 lines · 30 tokens per session scan A f1fcdf874242
reviewer is an agent published in the GitHub repository esneiderbravo/speclaw (1 stars, last pushed 7d ago), licensed MIT. It adds 37 tokens to every session and 692 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
integration-verifier
Read-only CP-4 gate. Cross-task wiring and global acceptance for multi-task specs. Cannot edit files.
code-reviewer
Expert code review specialist. Reviews code for quality, security, and maintainability.
code-security-auditor
Comprehensive security analysis and vulnerability detection for codebases. Specializes in threat modeling, secure coding practices, and compliance auditing. Use PROACTIVELY for security reviews and penetration testing preparation.
code-standards-enforcer
Enforce coding standards, style guides, and architectural patterns across projects. Specializes in linting configuration, code review automation, and team consistency. Use PROACTIVELY for code quality gates and CI/CD pipeline integration.
code-reviewer
Code review specialist that analyzes diffs and suggests improvements.
code-reviewer
Reviews the diff against the four execution principles. Use after changing code.