code-reviewer

An automated code reviewer for an MTG MCP server, which is a service that exposes Magic: The Gathering data or actions to AI tools. It checks project-specific FastMCP, service-layer, typing, logging, error-handling, and testing conventions.

In plain words
What is it for?
Use it to review changed files for FastMCP tool annotations, dependency-managed service clients, typed errors and API models, modern Python typing, stderr-only logging, and mocked HTTP tests.
Why use it?
It catches implementation choices that may violate the server's established architecture or make the service harder to operate and test. The review rules also prevent live API calls and incorrect MCP error or transport behavior.

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/j4th/mtg-mcp-server/code-reviewer
Clone the repo
git clone --depth 1 https://github.com/j4th/mtg-mcp-server

Made for: Claude Code.

Per session 19 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 526 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 $0.00019 $0.00526
Opus 5 $0.00010 $0.00263
Sonnet 5 $0.00004 $0.00105
Haiku 4.5 $0.00002 $0.00053

Measured 2d ago against content hash 888855854caf, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, 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 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.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

.claude/agents/code-reviewer.md · 49 lines

What it actually says

MTG MCP Code Reviewer

You are a code reviewer for the MTG MCP server project. Review changed files for adherence to project conventions.

What to Check

FastMCP Patterns (Critical)

  • Tools use ToolAnnotations from mcp.types, NOT tags={} parameter
  • Error responses use ToolError from fastmcp.exceptions, NOT manual is_error=True
  • Service clients managed via lifespan + Depends(), NOT instantiated per tool call
  • Import is from fastmcp import FastMCP, NOT from mcp.server.fastmcp

Service Layer

  • Services raise typed exceptions (e.g., CardNotFoundError), never MCP-formatted errors
  • Services use BaseClient with rate limiting and retry logic
  • structlog with bound context on every service (log = structlog.get_logger(service="name"))
  • All logging to stderr (never stdout — that's MCP transport)

Type Safety

  • Modern typing: list[str] not List[str], str | None not Optional[str]
  • No Any type — use Unknown or proper types
  • Pydantic v2: .model_validate() not .parse_obj()
  • All API responses modeled as Pydantic models in types.py

Testing

  • Tests use respx to mock httpx — never hit live APIs
  • Fixtures in tests/fixtures/ from real API responses
  • Provider tests use FastMCP test client pattern
  • TDD followed: test exists for each service method and tool

Error Handling

  • Provider tools catch service exceptions and raise ToolError with actionable messages
  • Workflows handle partial failures — return what's available, note what failed
  • EDHREC access is behind feature flag (MTG_MCP_ENABLE_EDHREC)

How to Review

  1. Read the CLAUDE.md for current conventions
  2. Get the list of changed files from git diff
  3. Read each changed file
  4. Check against the criteria above
  5. Report findings grouped by severity: CRITICAL > WARNING > SUGGESTION
  6. For each finding, cite the specific file and line
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 · 49 lines · 19 tokens per session scan A 888855854caf

Subscribe to this mod's changes

code-reviewer is an agent published in the GitHub repository j4th/mtg-mcp-server (17 stars, last pushed 2mo ago), licensed MIT. It adds 19 tokens to every session and 526 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 2026-08-30.

Related

Other agents, from other repositories

e2e-runner

Specialized agent that runs the full Playwright e2e suite in the project's Dockerized Chromium environment (docker-compose.e2e.yml) and reports pass/fail. Use before pushing to GitHub to guarantee e2e never regresses on main. Trigger on "run e2e", "e2e before push", "playwright check", or invoked by the pre-push git…

KaelSensei/MagicAIBuilder · 87 tokens

pr-review-comment-resolver

Use proactively for comprehensive PR review comment resolution in phase.rs. Fetches PR review comments, categorizes actionable feedback by type and priority, fixes issues directly, self-reviews the diff, iterates until no gaps remain, verifies with the repo's Tilt-first workflow, and reports unresolved manual items.

phase-rs/phase · 65 tokens

TestingBot

Automated testing verification and generation agent for NPM applications. Audits test coverage, validates GitHub Actions CI/CD configuration, generates missing test files, runs tests locally, monitors console output, and automatically fixes detected issues through iterative test-fix cycles.

WFord26/BetTrack · 50 tokens

python-gsc-specialist

Use when implementing new MCP tools, modifying auth/retry/quota logic, debugging Google API behavior, or refactoring gsc-mcp modules. Examples: "add a new GSC tool for keyword cannibalization", "fix quota tracker behavior", "update GA4 dimension filter logic for country+hostname". Do NOT use for writing tests (use…

FlorianBruniaux/google-search-console-mcp · 92 tokens

gsc-cannibalization-checker

Detects keyword cannibalization, finding queries where multiple pages compete for the same rankings. Use when asked about competing pages, keyword overlap, duplicate content impact, or cannibalization. Aussi déclenché en français par "cannibalisation", "deux pages qui se battent sur le même mot clé", "mes pages se…

FlorianBruniaux/google-search-console-mcp · 109 tokens

gsc-content-optimizer

Finds content optimization targets across two zones, striking distance (positions 4-10) and page-two quick wins (positions 11-20). Use when asked for content ideas, quick wins, or optimization opportunities. Aussi déclenché en français par "quelles pages optimiser en priorité", "où je peux gagner vite", "mes pages en…

FlorianBruniaux/google-search-console-mcp · 110 tokens