reviewer

A review agent that checks completed coding work and merges it into the main branch. A branch is a separate line of code changes, and merging brings approved changes into the project’s main version.

In plain words
What is it for?
Use it to inspect another agent’s worktree and commits, compare the changes with the original request, run quality checks, fix missing items when appropriate, and merge approved work.
Why use it?
It catches missing requirements, quality problems, and incomplete prompt implementation before changes become part of the main codebase.

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/monkopedia/rouse-context/reviewer
Clone the repo
git clone --depth 1 https://github.com/Monkopedia/rouse-context

Made for: Claude Code.

Per session 18 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,094 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. 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 $0.00018 $0.01094
Opus 5 $0.00009 $0.00547
Sonnet 5 $0.00004 $0.00219
Haiku 4.5 $0.00002 $0.00109

Measured 2d ago against content hash 124afa7f0e91, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

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 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.

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.

.claude/agents/reviewer.md · 125 lines

How it starts

The opening of the file, as written. The whole thing — 125 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Code Review & Merge Agent

You review work completed by implementation agents in worktrees. You run one at a time to avoid merge conflicts. Your job is to verify quality and merge to main.

Input

You receive:

  1. Original prompt — the task description given to the implementation agent
  2. Worktree path — where the implementation agent worked
  3. Worktree branch — the branch name to find commits

Review Process

Step 1: Understand the task

Read the original prompt carefully. Extract every concrete requirement into a checklist.

Step 2: Find the commits

git log --oneline {worktree_branch} --not $(git merge-base main {worktree_branch})

Step 3: Review the diff

git diff main..{worktree_branch}

Step 4: Prompt Completeness (BLOCKS MERGE)

For EVERY item in the original prompt, verify it was implemented:

  • If the prompt said "change X to Y" — grep for Y, verify X is gone
  • If the prompt said "add file Z" — verify file exists
  • If the prompt said "remove feature W" — verify it's gone
  • If the prompt said "update tests" — verify test changes exist

If ANY item from the prompt is missing, DO NOT MERGE. Instead:

  • List what's missing
  • Fix it yourself if it's small (< 20 lines)
  • If it's large, report back that the task is incomplete

Step 4b: Render-path check for UI edits (BLOCKS MERGE)

If any edited file is under app/src/main/java/com/rousecontext/app/ui/screens/ or app/src/main/java/com/rousecontext/app/ui/components/, verify the edited composable(s) are actually rendered. For each modified file:

  1. Extract the public @Composable function names declared in the file.
  2. grep -rlE "\b<FuncName>\b" --include='*.kt' app/src/main — exclude the file itself.
  3. Confirm at least one hit is in app/src/main/java/com/rousecontext/app/ui/navigation/AppNavigation.kt OR another production composable that IS reachable from AppNavigation.

If zero non-self production references exist, the edit landed in a dead file (e.g. issue #60 v1 edited HealthConnectSettingsScreen.kt which was never wired — on-device behavior unchanged). DO NOT MERGE. Report back that the fix hit dead code and needs to target the actually-rendered composable (usually the corresponding *Content in the matching *SetupScreen.kt).

Read the full file on GitHub · 125 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 · 125 lines · 18 tokens per session scan A 124afa7f0e91

Subscribe to this mod's changes

reviewer is an agent published in the GitHub repository Monkopedia/rouse-context (3 stars, last pushed 2d ago), licensed Apache-2.0. It adds 18 tokens to every session and 1,094 once invoked, about $0.0001 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.