agentleFS
YehudaFrankelYehudaFrankel/clankbrain/.claude/skills/end-session/SKILL.md

end-session

Skill16 starsChanged 7 months agoMITView on GitHub

Save session progress, extract lessons, update memory, push to GitHub if configured. Triggers on \"End Session\".

11 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 5 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 9 edited versions we read keep these.

  • 5

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Allowed Tools: Read, Edit, Write, Glob, Grep, Bash

    line 12
  2. Very specific:

    Check which memory layout exists: .claude/memory/lessons.md or tasks/lessons.md

    line 20
  3. Distinctive:

    If any runtime errors were solved this session, add them: Symptom | Root Cause | Fix

    line 24
  4. Distinctive:

    If any approach was tried and discarded this session, log it with why it was rejected

    line 28
  5. Distinctive:

    Add one-line summary of what changed

    line 38
  6. Very specific:

    If output files exist, update plan status to SHIPPED

    line 48
  7. Distinctive:

    If any planned task was completed this session, log estimated vs actual sessions

    line 51

“Common” is measured: kept by at least 6 in 10 of the 9 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Skill: end-session

Trigger: "End Session"

Description: Save session progress, extract lessons, update memory, push to GitHub if configured.

Allowed Tools: Read, Edit, Write, Glob, Grep, Bash


Steps

  1. Extract lessons (/learn):
    • Review conversation for bugs fixed, patterns discovered, decisions made
    • Check which memory layout exists: .claude/memory/lessons.md or tasks/lessons.md
    • Append lessons and decisions to whichever files exist
  2. Update error-lookup (tasks/error-lookup.md):
    • If any runtime errors were solved this session, add them: Symptom | Root Cause | Fix
    • Skip if no errors were encountered
  3. Update regret log (tasks/regret.md):
    • If any approach was tried and discarded this session, log it with why it was rejected
    • Skip if nothing was rejected
  4. Evolve check (/evolve-check if available):
    • Scan tasks/skill_scores.md for skills that needed corrections
    • Report which skills are healthy vs need attention
    • Skip if skill_scores.md is empty
  5. Update STATUS.md (project root):
    • Increment session number
    • Add one-line summary of what changed
  6. Update MEMORY.md (.claude/memory/MEMORY.md):
    • Set currentDate to today's date and session number
  7. Update skill_usage.md (tasks/skill_usage.md):
    • Log which skills fired this session with today's date
  8. Plan drift check:
    • Glob plans/ for files marked "Ready to Code" or "In Progress"
    • If output files exist, update plan status to SHIPPED
  9. Update velocity (tasks/velocity.md):
    • If any planned task was completed this session, log estimated vs actual sessions
    • Skip if no planned tasks were completed
  10. Push memory (if memory.ps1 exists in project root):
    powershell -NoProfile -ExecutionPolicy Bypass -File memory.ps1 push
    Skip silently if memory.ps1 doesn't exist.
  11. Report:
    Session complete.
    
    **Session N** -- [one-line summary]
    **Lessons extracted:** N
    - [one line per lesson]
    **Errors logged:** N (or "none")
    **Regrets logged:** N (or "none")
    **Files touched:** [list]

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.