# Save Learnings

> >

- **Type:** Skill
- **Install:** `agentstack add skill-enbiyagoral-claude-save-learnings`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [enbiyagoral](https://agentstack.voostack.com/s/enbiyagoral)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [enbiyagoral](https://github.com/enbiyagoral)
- **Source:** https://github.com/enbiyagoral/.claude/tree/main/.claude/skills/save-learnings

## Install

```sh
agentstack add skill-enbiyagoral-claude-save-learnings
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Save Session Learnings

## Purpose

Review the current session and persist any valuable learnings, corrections, patterns, or project knowledge to the memory system. This ensures future sessions benefit from today's discoveries.

## When to use

- End of a productive session with new discoveries
- After receiving user corrections or feedback
- After debugging a tricky issue
- When Stop hook reminds you to save learnings
- When user explicitly asks to save

## Process

1. **Read current memory state**
   - Read `MEMORY.md` index to see what already exists
   - Glob memory files to understand current coverage

2. **Identify what's new this session**
   - User corrections/feedback → `feedback` type memory
   - Project patterns discovered → `project` type memory
   - External resources/tools found → `reference` type memory
   - User preferences/role info → `user` type memory

3. **Check for updates vs new files**
   - If an existing memory covers the same topic, update it
   - Only create a new file if the topic is genuinely new
   - Keep memory file names descriptive: `feedback_testing.md`, `project_api_patterns.md`

4. **Write memory files**
   - Use frontmatter format: `name`, `description`, `type`
   - Keep content concise and actionable
   - Focus on "what" + "how to apply" — not verbose narratives

5. **Update MEMORY.md index**
   - Add one-line pointer for each new file
   - Keep entries under 150 chars
   - Keep total index under 200 lines

## Output

List of memory files created or updated, with a one-line summary of each.

## Quality bar

- Every saved memory has a clear "how to apply" — not just a fact dump
- No duplicate memories — always check existing files first
- MEMORY.md index stays clean and scannable
- Memory descriptions are specific enough to judge relevance in future sessions

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [enbiyagoral](https://github.com/enbiyagoral)
- **Source:** [enbiyagoral/.claude](https://github.com/enbiyagoral/.claude)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-enbiyagoral-claude-save-learnings
- Seller: https://agentstack.voostack.com/s/enbiyagoral
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.