memory-manage

Update CUDA.md memory and regenerate API.md and README.md documentation from repository code changes.

1|Updated May 12, 2026
One-click install
npx skills add https://github.com/tcytan/tecton --skill memory-manage-tcytan
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: memory-manage
Source: https://github.com/tcytan/tecton/tree/main/.claude/skills/memory-manage
Command: npx skills add https://github.com/tcytan/tecton --skill memory-manage-tcytan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill removes the friction of manually maintaining CLAUDE.md project memory and keeping API/README documentation in sync with code changes.

Core Features & Use Cases

  • CLAUDE.md maintenance: Update project memory with full, related (git-changed), or single-module handbook-style summaries.
  • Documentation generation: Generate or refresh API.md and README.md across the repository (including project-level docs) into .workflow/docs/.
  • Interactive routing & safety: Auto-detect intent, prompt when arguments are ambiguous, and verify output by ensuring only the expected target files are modified.

Quick Start

Ask the AI to run an incremental update and regenerate documentation for git-changed modules using the command: /memory-manage update-related --tool gemini

Frequently Asked Questions about memory-manage

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I keep CLAUDE.md project memory updated with recent git changes?

To keep CLAUDE.md project memory updated with recent git changes, use the incremental update mode to generate related module summaries based on your repository's git diff. This ensures your project memory reflects only the modified code structure.

How do I automatically generate API.md and README.md documentation from my repository code?

You can automatically generate API.md and README.md documentation from repository code by running the documentation generation feature. It refreshes project-level docs and module handbooks, outputting them directly into the .workflow/docs/ directory.

Can I use a fallback tool when the default AI model fails to update project documentation?

Yes, you can use a tool fallback mechanism with gemini, qwen, or codex to update project documentation. This ensures the documentation generation and CLAUDE.md maintenance processes complete successfully even if the primary model is unavailable.

What is the best way to manage full project memory refreshes and incremental documentation updates?

The best way to manage full project memory refreshes and incremental documentation is to use mode detection with interactive clarification. It supports full project refreshes, incremental related updates, and single-module deep documentation based on your specific request.

Does the documentation generation process verify that only expected target files are modified?

Yes, the documentation generation process includes safety verification of modified outputs. It ensures that only the expected target files, such as the generated API.md and README.md, are modified during the update workflow.

How does the system handle large scale repository documentation updates?

For large scale repository documentation updates, the system handles scale by using batching and agent execution. This allows it to process extensive code structures and generate comprehensive module handbooks without overwhelming the workflow.