End-of-Session Save
Review everything from this session and persist what's useful for future sessions.
Scope: every repo/context the session touched
- Route each learning to wherever it belongs, not just the primary working directory. Multi-repo and monorepo sessions are common — a fact learned here may belong in another repo's docs.
- Before concluding "nothing to save" for a given repo, check it for a now-stale doc or a convention this session introduced.
What to Save
- Conventions or patterns discovered during this session
- Commands, workflows, or setup steps that were non-obvious
- Gotchas or mistakes encountered (and how to avoid them)
- Reusable domain knowledge or specialized workflows
Where to Save (follow these rules strictly)
| Content type | Destination |
|---|---|
| Universal, prevents-mistakes-every-session rules | CLAUDE.md (append or update) |
| Specialized workflow or domain knowledge | A new or existing skill in .claude/skills/ |
| Reusable isolated assistant with scoped tools | .claude/agents/ |
| Deterministic automation (must always run) | .claude/hooks/ + settings.json |
| Nothing useful learned | Do nothing. Say so. |
Constraints
- Invoke
claude-md-and-folder-standardsas the single source of truth for placement and Claude-Code config rules — don't restate them here (an inlined copy drifts). When a learning becomes a new skill, also invokewriting-skillsfor how to write the SKILL.md. - Highlights only: CLAUDE.md ≤200 lines; skills stay concise (≤500); add only what Claude would get WRONG without; no duplicating existing files; terse imperative bullets.
- Repo-scoped. This saves to a repo's
.claude/(works in Claude Code and in Cowork/cloud sessions on a cloned repo). It can't touch the Claude app's account-level personal skills — those are managed in the app's UI, not by writing a file.
Do NOT
- Write to MEMORY.md
- Add content that's already captured elsewhere in .claude files
- Add content with no clear future value
- Exceed the 200-line CLAUDE.md cap