Plugin Claude Code
A foundation for agents that keeps three levels of memory: raw events, reviewed summaries, and longer-term knowledge. It also describes approval gates and a TDD workflow, meaning tests are written before the code.
Plugin Claude Code
A project structure for keeping an AI agent’s memory outside the model in three layers: persistent notes, an index, and detailed records. It also provides templates, approval controls, event logs, context-saving procedures, and weekly review tools.
Hook
Runs when a session starts, executing on-session-start.sh via bash. From reatcat/l123-harness.
Hook
Runs before the agent uses a tool for Edit, Write, MultiEdit, NotebookEdit, edit, write and str_replace_editor tool calls, executing gate-guard.sh via bash. From reatcat/l123-harness.
Hook
Runs after a tool call finishes, executing on-post-tool.sh via bash. From reatcat/l123-harness.
Hook
Runs before the context is compacted, executing on-pre-compact.sh via bash. From reatcat/l123-harness.
Hook
Runs when the agent finishes a response, executing on-stop.sh via bash. From reatcat/l123-harness.
Hook
Runs when a session ends, executing on-session-end.sh via bash. From reatcat/l123-harness.
Skill Claude CodeCodex
A weekly review process for event logs and leftover task records, turning repeated problems or useful lessons into proposed memory, knowledge, or skill entries.
Skill Claude CodeCodex
An installer for a three-level project memory structure: a main instruction file, an index, and detailed notes in an `.agents` directory. It also adds templates and protection settings.
Skill Claude CodeCodex
Test-driven development with red-green-refactor loop. Use when user wants to build features or fix bugs using TDD, mentions "red-green-refactor", wants integration tests, or asks for test-first development.
Skill Claude CodeCodex
Create new agent skills with proper structure, progressive disclosure, and bundled resources. Use when user wants to create, write, or build a new skill.