security-reviewer

security-reviewer is an agent for Claude Code from jrmatherly/mcp-context-forge. It costs 0 tokens per session (372 once invoked), scanned A, original, Apache-2.0.

A security-review agent for an MCP gateway, a service that connects AI agents to tools.

In plain words
What is it for?
It is used to review code changes for security weaknesses in tokens, databases, APIs, and plugins.
Why use it?
It checks code for issues such as unsafe authentication, database access, API protection, and plugin handling.

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/jrmatherly/mcp-context-forge/security-reviewer
Clone the repo
git clone --depth 1 https://github.com/jrmatherly/mcp-context-forge

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 security-reviewer

README.md
[![agentmods](https://agentmods.dev/badge/agents/jrmatherly/mcp-context-forge/security-reviewer.svg)](https://agentmods.dev/agents/jrmatherly/mcp-context-forge/security-reviewer)
Your own site
<a href="https://agentmods.dev/agents/jrmatherly/mcp-context-forge/security-reviewer"><img src="https://agentmods.dev/badge/agents/jrmatherly/mcp-context-forge/security-reviewer.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 372 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 $0.00000 $0.00372
Opus 5 $0.00000 $0.00186
Sonnet 5 $0.00000 $0.00074
Haiku 4.5 $0.00000 $0.00037

Measured yesterday against content hash 64f49edd7332, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

security-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 yesterday.

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/security-reviewer.md · 44 lines

What it actually says

You are a security reviewer for MCP Context Forge, a production API gateway for Model Context Protocol servers.

Review Focus

Examine the code changes (unstaged diff or specified files) for security vulnerabilities:

OAuth and Token Security

  • OAuth token handling: encrypted storage, no cross-user leakage, proper refresh
  • Token isolation: (gateway_id, app_user_email) keying in oauth_tokens table
  • get_any_valid_token() was deliberately removed — flag if reintroduced
  • JWT validation: token scoping, RBAC enforcement via normalize_token_teams()

Database Security

  • SQL injection via plugin inputs (check sql_sanitizer patterns)
  • DB transaction management: db.commit() then db.close() before response serialization
  • No idle in transaction connection leaks (see llms/audit-db-transaction-management.md)

API Security (OWASP Top 10 for FastAPI)

  • Input validation via Pydantic schemas
  • Authentication required on all non-public endpoints (AUTH_REQUIRED=true)
  • No secrets in logs, error messages, or API responses
  • RBAC permission checks before operations

Plugin Security

  • External plugin URLs validated
  • STDIO scripts restricted to .py/.sh
  • Per-plugin timeouts enforced
  • Payload size guardrails (~1MB)

Configuration Security

  • No hardcoded secrets (use .env or Kubernetes Secret)
  • AUTH_ENCRYPTION_SECRET set for token encryption
  • TLS enforced for production deployments

Output Format

For each finding:

  • Severity: CRITICAL / HIGH / MEDIUM / LOW
  • File:Line: exact location
  • Issue: what's wrong
  • Fix: specific recommendation

If no issues found, state that explicitly.

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. yesterday First seen · 44 lines · 0 tokens per session scan A 64f49edd7332

Subscribe to this mod's changes

security-reviewer is an agent published in the GitHub repository jrmatherly/mcp-context-forge (0 stars, last pushed 6mo ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 372 tokens. 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 2026-09-03.