From android
Review code changes in an Android project (app or library) using the Codex MCP server. Discovers project context at runtime.
How this skill is triggered — by the user, by Claude, or both
Slash command
/android:codex-reviewThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Review the current changes in the repository using Codex for in-depth analysis.
Review the current changes in the repository using Codex for in-depth analysis.
Follow the Common Command Instructions from review-base.md, with these Codex-specific customizations.
Execute steps 1–3 from the Common Command Instructions section in the template:
Read the guidelines from review-base.md and follow Step 4 (Generate Customized Prompt), using this Codex-specific introduction template. Substitute discovered project values — do NOT leave placeholders literal, and do NOT invent project context you did not actually find.
Codex-Specific Introduction Template:
You are an expert code reviewer analyzing code changes for [PROJECT_NAME] ([PROJECT_TYPE]).
You are running in Codex MCP with workspace-write sandbox mode, which gives you access to:
- Bash commands (git, file operations, etc.)
- File reading capabilities
- Full repository access
Project context:
- Working directory: [PROJECT_PATH]
- Current branch: [BRANCH_NAME]
- Main branch: [MAIN_BRANCH]
- Architecture: [ARCHITECTURE_SUMMARY] (omit this line if not discovered)
Your task is to perform a comprehensive code review covering workspace changes
and branch context as needed. Your separate execution environment allows for
thorough analysis and extensive git operations.
Then complete the prompt by:
[AVAILABLE_READ_TOOL] with "bash cat command or file reading" in git analysis instructionsCall the mcp__codex__codex tool with the fully customized prompt:
Tool: mcp__codex__codex
Parameters:
{
"cwd": "<absolute-path-to-project-root>",
"sandbox": "workspace-write",
"prompt": "<fully customized prompt from step 4>"
}
Present the results directly to the user with clear sections as specified in the template's Expected Output Format.
Simply type /codex-review in Claude Code to trigger this review workflow.
/codex-review: Uses Codex MCP server (separate execution environment)
/agent-review: Uses Claude Code's Task agent (subagent within current session)
Both use the same review criteria from review-base.md.
Provides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub yshrsmz/cc-plugins --plugin android