Session end

Skill itsjwill/claude-memory/skills/session-end

Persistent memory for Claude Code. Auto-capture decisions, learnings, client info. Cloud backup to Supabase. Never lose context again. Free alternative to Mem.ai, Rewind AI, Personal AI.

Install

npx -y skills add itsjwill/claude-memory --skill session-end

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

14 stars14 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Generate and save a session summary before ending conversation. IMPORTANT: Use this PROACTIVELY when: - User says "goodbye", "that's all", "thanks", "done for now" - Session has been productive with decisions/learnings - Conversation is naturally winding down DO NOT ask permission - just summarize and save silently.

SKILL.md

2.7 KB, as published. Nobody here has run it

Session End Summary

Generate a concise summary of this session and save it to memory.

When to Invoke

Proactively invoke this skill when:

User indicates session is ending ("bye", "thanks", "done", "that's all")
Long productive session is wrapping up
Significant work was completed

Summary Format

Generate a summary with:

One-line summary: What was the main focus of this session?
Key outcomes: What was accomplished? (2-4 bullet points)
Decisions made: Any choices or directions decided
Open items: What's left to do or follow up on
Tags: Project names, technologies, topics discussed

Execution Steps

Review conversation: Identify main topics, decisions, outcomes
Generate summary: Create concise session summary
Extract tags: Pull project names, technologies, client names
Store to memory: Use memory_store with type "progress"
Update MEMORY.md: If significant, append to quick reference
Silent completion: Do NOT announce unless user explicitly invoked /session-end

Storage Format

{
  "content": "Session summary: [one-line description]\n\nOutcomes:\n- [outcome 1]\n- [outcome 2]\n\nDecisions:\n- [decision 1]\n\nOpen items:\n- [item 1]",
  "metadata": {
    "type": "progress",
    "tags": "session-summary,[project],[topics]",
    "source": "session-end-skill"
  }
}

Example

Session about trading bot optimization:

Session summary: Optimized Billy V4 bot parameters based on backtest results

Outcomes:
- Disabled shorts (losing $2,226 in backtests)
- Added funding rate filter (<0.02%)
- Tightened first scale to +2%

Decisions:
- Billy stays LONGS ONLY
- Blood keeps bidirectional with asymmetric Z

Open items:
- Monitor live performance for 1 week
- Consider adding SOL to tradeable coins

Tags: session-summary, trading, billy-v4, optimization, backtest

MEMORY.md Update

If session produced significant outcomes, append to MEMORY.md under relevant section:

### [Date] - [Project/Topic]
- [Key outcome or decision in one line]

Only update MEMORY.md for:

New client information
Major architectural decisions
Important file path references
Critical learnings that should be instantly visible