claude-code-recipes

A set of eight Claude Code plugins for reviewing plans, checking shell commands, and adding project context to helper agents.

In plain words
What is it for?
It helps review implementation plans, block destructive shell operations, and provide project files, rules, Git status, and structure to planning or exploration agents.
Why use it?
It reduces the risk of approving flawed plans or running dangerous commands. It also helps helper agents understand the project before they begin.

Plugin 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.

Claude Code
/plugin marketplace add bradlay/claude-code-recipes
agentmods
npx agentmods add plugins/bradlay/claude-code-recipes/marketplace
Clone the repo
git clone --depth 1 https://github.com/bradlay/claude-code-recipes

Made for: Claude Code.

Per session not measured What this adds to a session before it is invoked.
When invoked not measured Not applicable: nothing here is loaded into a session.
Security scan A 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Security

Grade A, and why

claude-code-recipes 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 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

"description": "PreToolUse(Bash) hook that evaluates every shell command against a configurable rule set. Blocks destructive git ops, filesystem-nuking commands, untrusted curl-pipe-to-shell patterns, and identity-manage
.claude-plugin/marketplace.json · 50 lines

How it starts

The opening of the file, as written. The whole thing — 50 lines — stays where its author put it; the contents beside it link to each section on GitHub.

{
  "name": "claude-code-recipes",
  "owner": {
    "name": "bradlay",
    "url": "https://github.com/bradlay"
  },
  "plugins": [
    {
      "name": "plan-review-loop",
      "source": "./plugins/plan-review-loop",
      "description": "Closed-loop plan review on ExitPlanMode. Pick a backend per session (Opus 4.8, Sonnet 5, codex, or Gemini 3.1 Pro via agy) — only probe-verified ones are offered; autoswe reviews locally. Denies plan exit on P0/P1 findings; loop closes when the plan is clean."
    },
    {
      "name": "bash-guard",
      "source": "./plugins/bash-guard",
      "description": "PreToolUse(Bash) hook that evaluates every shell command against a configurable rule set. Blocks destructive git ops, filesystem-nuking commands, untrusted curl-pipe-to-shell patterns, and identity-management changes. Rules are user-editable."
    },
    {
      "name": "subagent-context-injector",
      "source": "./plugins/subagent-context-injector",
      "description": "SubagentStart hook that injects project context (CLAUDE.md, .claude/rules/*.md, git state, top-level structure) into Plan and Explore subagents so they boot with full project awareness instead of cold."
    },
    {
      "name": "precompact-context-keeper",
      "source": "./plugins/precompact-context-keeper",
      "description": "PreCompact hook that injects CLAUDE.md + a tiny git-state summary as a systemMessage just before compaction so the post-compaction model still has the project framing."
    },
    {
      "name": "posttooluse-bash-audit",
      "source": "./plugins/posttooluse-bash-audit",
      "description": "PostToolUse(Bash) hook that appends every executed Bash command (timestamp + session id + 200-char summary) to a local audit log. Stays local, fail-open."
    },
    {
      "name": "branch-warn",
      "source": "./plugins/branch-warn",
      "description": "UserPromptSubmit hook that emits 'on branch X' once per hour as a systemMessage. Catches the 'wait, I should be on a feature bra

Read the full file on GitHub · 50 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. yesterday First seen · 50 lines scan A c11c5b4d84a0

Subscribe to this mod's changes

claude-code-recipes is a plugin published in the GitHub repository bradlay/claude-code-recipes (2 stars, last pushed 10d ago), licensed MIT. Its token cost is not measured: this kind of file is read by the harness, not the model. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.