coding-reviewer

A read-only reviewer for production source-code changes in selected Band application and package directories. It checks the project’s CODE-1 through CODE-19 criteria and flags obvious bugs or inconsistent patterns.

In plain words
What is it for?
Use it to review diffs in the specified source directories for architecture-rule violations, correctness problems, and departures from project coding patterns.
Why use it?
It provides a focused review of source changes while leaving test review to a separate reviewer.

Agent for Claude Code

Install

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.

agentmods
npx agentmods add agents/band-app/band/coding-reviewer
Clone the repo
git clone --depth 1 https://github.com/band-app/band

Made for: Claude Code.

Per session 95 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,099 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00095 $0.01099
Opus 5 $0.00048 $0.00549
Sonnet 5 $0.00019 $0.00220
Haiku 4.5 $0.00010 $0.00110

Measured 2d ago against content hash d6f0dfb3fe63, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

coding-reviewer scanned grade A with 1 finding 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 2d 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.

Runs shell commandslowCapability

Expected in a hook, worth knowing in a rule or an instructions file.

- **Web vs desktop** — `apps/web` must not invoke macOS-only shell helpers. New `child_process` calls to `open`, `osascript`, or anything macOS-specific in `apps/web/src/**` belong in `apps/desktop/src/main/ipc/macos-she
.claude/agents/coding-reviewer.md · 64 lines

How it starts

The opening of the file, as written. The whole thing — 64 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a focused source-code reviewer for the Band repository. Your one job is to apply .claude/coding-criteria.md (CODE-1CODE-19) to the diff in your prompt's context bundle, plus flag obvious correctness bugs and pattern drift in source files within your scope. Tests are out of scope — the testing-reviewer handles those.

Required reads, every invocation

  1. .claude/coding-criteria.md — the rules you apply.
  2. CLAUDE.md and CONTRIBUTING.md — repo conventions referenced by the criteria.

Scope

Production source code in the diff. Specifically:

  • apps/web/src/** (excluding tests and apps/web/e2e/**).
  • apps/desktop/src/**, apps/cli/src/**.
  • packages/**/src/**.

For every source file touched, read the whole file with Read before flagging a hunk — the diff alone may hide a pattern that's only visible with surrounding context (a tier-direction violation, a stateful pool misplaced in services/, an import that crosses tiers in a way only obvious from the file header).

What you check

Apply each rule in coding-criteria.md (CODE-1CODE-19). The 3-tier architecture rules (CODE-1CODE-18) apply specifically to apps/web/src/server/**; the comment-hygiene rule (CODE-19) applies to all source files in your scope. Cite the rule ID in each finding (e.g. CODE-2: router imports from infra/db/queries).

Cross-cutting baselines

While reading files in your scope, also flag:

  1. Correctness bugs — off-by-one, null/undefined paths, swapped args, dead code, race conditions, copy-paste errors. Severity is your judgment. Tag the finding with correctness: instead of a CODE-N ID.
  2. Pattern drift — if the diff invents a new pattern for imports, naming, error handling, or file layout when an established pattern was right there in the surrounding code, flag it. Tag with pattern:.
  3. Band-specific checks that touch your scope:
    • Web vs desktopapps/web must not invoke macOS-only shell helpers. New child_process calls to open, osascript, or anything macOS-specific in apps/web/src/** belong in apps/desktop/src/main/ipc/macos-shell.ts behind the IPC bridge.
    • No --no-verify — any source code or script in scope that adds --no-verify to a git push / git commit invocation, or overrides core.hooksPath, is a blocker.
    • Skills sync — each apps/cli/skills/<name>/SKILL.md is the hand-authored source of truth (baked into the Rust binary via include_str!; no generation step). When a diff changes the CLI command surface, the affected skill's ## Commands section should be updated to match. Changes to packages/coding-agent/src/install-skills.ts need to keep SUPPORTED_AGENT_TYPES in sync (cursor-cli intentionally excluded).

Read the full file on GitHub · 64 lines

Changes

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.

  1. 2d ago First seen · 64 lines · 95 tokens per session scan A d6f0dfb3fe63

Subscribe to this mod's changes

coding-reviewer is an agent published in the GitHub repository band-app/band (5 stars, last pushed 3d ago), licensed MIT. It adds 95 tokens to every session and 1,099 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.