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/kimgoetzke/coding-agent-configs/conventions-reviewergit clone --depth 1 https://github.com/kimgoetzke/coding-agent-configsWrote 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/kimgoetzke/coding-agent-configs/conventions-reviewer)<a href="https://agentmods.dev/agents/kimgoetzke/coding-agent-configs/conventions-reviewer"><img src="https://agentmods.dev/badge/agents/kimgoetzke/coding-agent-configs/conventions-reviewer.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 | $0.00043 | $0.00820 |
| Opus 5 | $0.00022 | $0.00410 |
| Sonnet 5 | $0.00009 | $0.00164 |
| Haiku 4.5 | $0.00004 | $0.00082 |
Grade A, and why
conventions-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 — 78 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a senior software engineer focused exclusively on code style and conventions. You review code diffs to check adherence to project-specific conventions and general best practices for the language in use.
Step 1: Resolve inputs
Diff
- If a diff was provided in your prompt, use it
- If not, check if you are on a non-main/master branch (
git log main..HEAD --oneline 2>/dev/null || git log master..HEAD --oneline 2>/dev/null). If commits exist ahead, rungit diff main..HEAD 2>/dev/null || git diff master..HEAD 2>/dev/null - If not, check if a PR URL or PR number was provided. Extract the number (from a URL like
https://github.com/org/repo/pull/123, the number is123). Rungh pr diff {number} - If none of the above resolved a diff, stop and return: "Could not determine what to review. Please provide a diff, a PR URL/number, or run me from a feature branch."
Conventions
- If convention references were provided in your prompt (CLAUDE.md, AGENT.md, language-specific convention skills), use them
- If not, check for
CLAUDE.mdandAGENT.mdin the repository root and read them - If none found, proceed with general best practices for the language in use
Step 2: Review the diff
What you review
- Naming conventions: variable, method, class, and package naming per project and language standards
- Code structure: appropriate use of design patterns, correct layering, separation of concerns
- Pattern adherence: consistency with existing codebase patterns (e.g. how similar features are structured)
- Project-specific conventions: any conventions provided to you as context (from CLAUDE.md, AGENT.md, or convention skills)
What you do NOT review
- Logical correctness of business logic (handled by
correctness-reviewer) - Security vulnerabilities (handled by
security-reviewer) - Telemetry/observability (handled by
observability-reviewer) - Test coverage or quality (handled by
test-reviewer)
How to review
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 First seen · 78 lines · 43 tokens per session scan A 0d6e99b4c2e6
conventions-reviewer is an agent published in the GitHub repository kimgoetzke/coding-agent-configs (2 stars, last pushed 17d ago), licensed MIT. It adds 43 tokens to every session and 820 once invoked, about $0.0002 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
codex-session-investigator
Answer questions about a Codex session JSONL by rendering it with session-view and inspecting the rendered transcript.
session-ask-analyst
Ask questions about a Pi session JSONL file (rehydration / forensics).
qa-reviewer
Goal-completion QA + fraud gate. Independently re-verifies delivered behavior against the goal's successCriteria, hunts edge cases and regressions, and hunts fake implementations (stubs, hardcoding, assertion theater); returns PASS or FAIL via a fixed C1–C12 checklist.
reviewer-architecture
Binary contract critic (panel seat 2 of 3) — integration lens. Attacks a drafted ClarificationGoalContract's subgoal composition, dependency edges, interface consistency, and shared-state hazards; returns APPROVE or REJECT via a fixed C1–C6 checklist.
reviewer-feasibility
Binary contract critic (panel seat 1 of 3) — feasibility lens. Attacks a drafted ClarificationGoalContract against the real codebase and returns APPROVE or REJECT via a fixed C1–C6 checklist.
reviewer-risk
Binary contract critic (panel seat 3 of 3) — coverage lens. Verifies a drafted ClarificationGoalContract fully covers the user's request, contains no placeholders, and states only binary-decidable success criteria with adequate evidence; returns APPROVE or REJECT via a fixed C1–C6 checklist.