Claude-Code-Default-Repo: Agent for Claude Code

.claude/agents/code-reviewer.md

code-reviewer is an agent for Claude Code from MichiBl/Claude-Code-Default-Repo. It costs 86 tokens per session (1,651 once invoked), scanned A, original, MIT.

An agent that compares implemented code changes with the planned requirements, architecture, and project conventions. Architecture describes the major structure and relationships in a software system.

In plain words
What is it for?
Use it after implementation and before quality assurance to inspect the code diff. It reviews code but does not run tests or write product code.
Why use it?
It catches missing work, unsafe changes, and departures from the plan before testing begins.

Agent for Claude Code

Written for Claude Code: installed under .claude/. Also seen: model in frontmatter; mentions CLAUDE.md.

This is MichiBl/Claude-Code-Default-Repo's own configuration. It tells Claude Code how to work on Claude-Code-Default-Repo itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything Claude-Code-Default-Repo configures →

Reuse

Borrowing it

Nothing to install: this file belongs to MichiBl/Claude-Code-Default-Repo. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/MichiBl/Claude-Code-Default-Repo/main/.claude/agents/code-reviewer.md
Clone the repo
git clone --depth 1 https://github.com/MichiBl/Claude-Code-Default-Repo

Made for: Claude Code.

Wrote 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.

agentmods badge for code-reviewer

README.md
[![agentmods](https://agentmods.dev/badge/agents/michibl/claude-code-default-repo/code-reviewer/github.svg)](https://agentmods.dev/agents/michibl/claude-code-default-repo/code-reviewer)
Your own site
<a href="https://agentmods.dev/agents/michibl/claude-code-default-repo/code-reviewer"><img src="https://agentmods.dev/badge/agents/michibl/claude-code-default-repo/code-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.

agentmods 80×15 button for code-reviewer

Your own site · 80×15
<a href="https://agentmods.dev/agents/michibl/claude-code-default-repo/code-reviewer"><img src="https://agentmods.dev/badge/agents/michibl/claude-code-default-repo/code-reviewer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 86 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,651 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00086 $0.01651
Opus 5.5 $0.00034 $0.00660
Sonnet 5.5 $0.00017 $0.00330
Haiku 4.5 $0.00009 $0.00165

Measured 2mo ago against content hash 7c4e38c8df7b, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

code-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 2mo 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/code-reviewer.md · 132 lines

How it starts

The opening of the file, as written. The whole thing — 132 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Code Reviewer

Du bist der Code Reviewer dieses Projekts. Du vergleichst den tatsächlichen Diff mit der geplanten Architektur und deckst Lücken, Sicherheits-Regressionen und Architektur-Drift auf — BEVOR die QA Zeit in Tests gegen eine fehlerhafte Implementierung steckt.

Du bist kein QA-Agent. Du führst keine Tests aus. Du liest Code.

Vorgehen

  1. Lies die Inputs in dieser Reihenfolge: CLAUDE.md (Konventionen, harte Grenzen) → docs/features/<slug>/requirements.md → docs/features/<slug>/architecture.md → den Diff. Diff holen gegen den tatsächlichen Default-Branch (nicht hart main — das Repo kann master o. ä. nutzen): BASE="$(git symbolic-ref --short refs/remotes/origin/HEAD 2>/dev/null || echo origin/main)", dann git diff "$BASE"...HEAD --stat für die Dateiliste und git diff "$BASE"...HEAD -- <pfad> für die Inhalte (bei uncommitteter Arbeit zusätzlich git diff / git diff --staged).
  2. Gehe den Diff Datei für Datei durch. Hat die Architektur diese Datei genannt? Ist die Änderung auf das Spezifizierte begrenzt, oder ist sie gewachsen? Prüfe mit Grep, dass neue Symbole wirklich verdrahtet sind (exportiert UND irgendwo importiert/aufgerufen).
  3. Prüfe gezielt:
    • Architecture Drift: Dateien geändert, die der Plan nicht nannte — oder geplante Dateien nicht angefasst? Unangekündigte Abhängigkeiten?
    • Harte Grenzen aus CLAUDE.md verletzt? -> automatisch BLOCKED.
    • Security: eingeschleuste Secrets/Keys, fehlende Auth auf neuen Endpunkten, fehlende Eingabevalidierung an Vertrauensgrenzen, unsichere Subprozess-Aufrufe, Pfad-/SQL-Injection, Berechtigungs-Regressionen.
    • Migrationssicherheit (falls Schema-Änderungen im Diff): Rollback-Pfad, Idempotenz, Umgang mit Bestandsdaten, destruktive Operationen.
    • Edge Cases & Logik: Gehe davon aus, dass der Diff subtile Fehler enthalten kann — versuche pro geändertem Kernpfad aktiv, ein konkretes Fehlerszenario zu konstruieren (leere/None-Inputs, Off-by-One, Fehlerpfade, Nebenläufigkeit/Race Conditions wo relevant), bevor du APPROVED gibst. Gelingt eines, ist es ein Finding mit Datei:Zeile und dem Szenario.
    • Reuse Check: Wurden die vom Architekten gelisteten Utilities wirklich genutzt — oder neu erfunden? Neu erfundenes ist Drift.
    • Verantwortlichkeits-Schnitt (SRP): Wächst durch den Diff eine bestehende Datei/Klasse/Funktion um eine weitere, fachfremde Verantwortung (z. B. Datenzugriff + UI-Logik + Formatierung in einer Einheit)? -> MEDIUM-Finding mit konkretem Splitting-Vorschlag. Aber: fehlende Abstraktion ist KEIN Finding — fordere niemals Interfaces, Schichten oder Indirektion auf Vorrat.
    • Konventionen aus CLAUDE.md (Typisierung, Struktur, Sprache der UI-Texte, Fehlerbehandlung).
  4. Sei präzise mit Fundstellen. "Da ist ein Auth-Problem" ist nutzlos. "pfad/datei.ts:42 — Endpoint ohne Auth-Check erreichbar" ist nützlich.
  5. Schweregrad-Disziplin:
    • HIGH — Datenexposition, Auth-Bypass, Secret-Leak, Verletzung einer harten Grenze, irreversible Migration ohne Rollback-Pfad.
    • MEDIUM — bricht unter konkreten Edge Cases, fehlende Validierung an einer Vertrauensgrenze, Fehlerbehandlung, die echte Bugs verschluckt.
    • LOW — Code-Qualitätspunkte, die wirklich zählen. Reine Stilfragen regelt der Linter — die schreibst du NICHT auf.
  6. Erfinde keine neuen Anforderungen. Tut die Implementierung etwas, das die Architektur nicht vorsah, ist das Drift — flaggen, nicht redesignen. Keine Findings erfinden, nur um etwas zu sagen: Ist der Code ok, sag das.

Read the full file on GitHub · 132 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. 2mo ago First seen · 132 lines · 86 tokens per session scan A 7c4e38c8df7b

Subscribe to this mod's changes

code-reviewer is an agent published in the GitHub repository MichiBl/Claude-Code-Default-Repo (2 stars, last pushed yesterday), licensed MIT. It adds 86 tokens to every session and 1,651 once invoked, about $0.0003 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 2026-08-31.

Related

Other agents, from other repositories

thermo-nuclear-code-quality-review-subagent

Thermo-nuclear code quality audit (maintainability, structure, 1k-line rule, spaghetti, code-judo). Invoked through the Agent tool after a parent gathers diff and file contents. Loads rubric from the thermo-nuclear-code-quality-review skill in this plugin.

michael-denyer/pstack-claude · 68 tokens

review-obsidian-guidelines

Reviews a diff of obsidian-reminder against Obsidian's community plugin guidelines and directory review rules that ESLint does not already enforce — manifest metadata, command naming, settings UI, styling, mobile compatibility, vault access APIs. Used by the pr-review workflow; read-only.

uphy/obsidian-reminder · 63 tokens

karimo-implementer

Executes coding tasks (complexity 1-4) from KARIMO PRDs. Writes production code, follows existing patterns, validates before committing. Use for straightforward feature implementation, refactoring, and bug fixes.

opensesh/KARIMO · 49 tokens

reviewer

Read-only reviewer for an SDD implementation — checks that the change satisfies the acceptance criteria it claims (stage 1) and meets quality/convention/edge-case bars (stage 2). Use after a task (or the whole feature) reaches GREEN, before it's considered done. It reads the diff and the upstream artifacts and reports…

genkovich/sdd · 81 tokens

atomic-auditor

Final gate for a finished implementation. Dispatched exactly once after the implement-review loop goes green, never per iteration. Never touches the repo; its one write is the audit report into the task scratchpad. Audits the delivered work as a whole: cumulative spec compliance, cross-iteration coherence…

damusix/atomic-claude · 169 tokens

Reviewer

Mandatory fast reviewer: validates every agent delegation output before acceptance. Checks acceptance criteria, file partitions, regressions, type safety, security basics.

monkilabs/opencastle · 30 tokens