Skill Claude CodeCodex
A writing guide for creating or revising AiWiki entries about misconceptions. AiWiki is a documentation project, and these entries are written in Markdown with extra page metadata.
Skill Claude CodeCodex
A specialized writing workflow for AiWiki entries about protecting privacy when using AI. It follows a required article structure, privacy-specific metadata, first-person safety rules, review gates, and reliable original sources.
Skill Claude CodeCodex
A recurring research workflow for AiWiki articles about protecting privacy when using large language models, which are systems that generate text from prompts. It finds reliable original sources, removes duplicate topics, applies three quality checks, and sends accepted items through the privacy-article process.
Skill Claude CodeCodex
A scheduled research and publishing workflow for AiWiki, a knowledge site. It finds reliable sources about common mistakes when using AI, removes duplicate topics, checks quality, and submits new entries through the site’s review steps.
Skill Claude CodeCodex
A translation guide for making English copies of AiWiki misconception entries. It places the translated Markdown files in the matching i18n/en/ folder.
Instructions file
Project memory and writing instructions for AIWiki, including the required reading order and rules for keeping content and style consistent across sessions. It also directs agents to use dedicated skills for writing and translation.
Plugin Claude Code
Plugin marketplace listing 1 plugin: agent-guard.
Plugin Claude Code
Outbound change control for AI coding agents. Gates git push, npm publish, docker push, gh release, non-local HTTP mutations, and rm -rf before they happen, against a bundled policy, with Ed25519-signed audit receipts. Local-first: no cloud, no telemetry.
Hook Claude Code
Runs before the agent uses a tool for Bash tool calls, executing pre-commit-review.sh via bash. From XuebinMa/agent-guard.
Hook Claude Code
Runs after a tool call finishes for Edit and Write tool calls, executing post-edit-review.sh via bash. From XuebinMa/agent-guard.
Settings file Claude Code
Agent settings declaring 2 hook events (PreToolUse, PostToolUse).
Instructions file CodexOpenCode
Instructions for XuebinMa/agent-guard: This repository keeps a single source of truth for agent and contributor guidance in CLAUDE.md. Read it before working in this repo — it covers the project overview, build & test commands, workspace architecture, key types, the core execution pipeline, and the working rules.
Instructions file
Instructions for XuebinMa/agent-guard, covering claude.md, project overview, build & test commands, canonical local verification entrypoint and rust-only verification path.