메뉴
Save or update the current session. Use this skill when asked to handoff, save session, update session, start a new chat, etc. Always call this before using the `update_session` MCP tool directly.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Generate or update CHANGELOG.md from git commits using semantic versioning. Use when the user says "/changelog", wants to update the changelog, or asks to generate release notes.
Find typos, grammar errors, punctuation/capitalization mistakes, and banned-phrase voice slips in prose and code comments. Use when the user says "/typo-check", wants to scan for typos, or before shipping user-facing text. Reports findings and confirms before fixing.
Commit, docs, changelog, and push in one go. Use when the user says "/ship", wants to ship their changes, or asks to commit and push everything.
Import conversation history from AI providers (Claude, Copilot, Codex, Cursor) and generate stories from the imported data. Use on first session to bootstrap memory from existing conversations, or anytime the person wants to import history from another provider/machine.
Research a topic and save findings to Charlieverse knowledge. Use whenever the user wants to research something, look up documentation, investigate a library or tool, explore a concept, or asks "what do you know about X". Also trigger when the user says "/research", wants to build up knowledge about a subject, or asks you to "look into" something — even if they don't use the word "research" explicitly.
Scan commits and diffs for decisions, record them as immutable ADRs in docs/decisions/.
| name | session-save |
| description | Save or update the current session. Use this skill when asked to handoff, save session, update session, start a new chat, etc. Always call this before using the `update_session` MCP tool directly. |
If skipping a step, no need to explain that you did, just skip it.
Call update_session with:
what_happened:
State at end of session, single paragraph. What's true NOW — what got resolved, what's still open, what's blocking the next pickup. One line of texture (grind, breakthrough, spiral) if relevant. Anchor to artifacts — "3 tests in test_banned_words.py fail because X" beats "some tests broke."
If it's longer than 4 sentences, it will be truncated. So be conscise
IMPORTANT DO NOT INCLUDE:
- Anything `git log` knows (diffs, file lists, commit messages).
- Anything that belongs in a memory (preferences, decisions, moments). Don't duplicate.
- Process narration ("we discussed," "we considered," "we decided to").
- Chronology ("first we did X, then Y, then Z").
- Long file paths (use relative file paths)
- Any information that already included in a memory
- Number of files/lines changed
for_next_session: Bulleted TODOs, action-verb first, one line each. Each item should be something the next Charlie can pick up and do without re-reading the transcript. "Continue banned-words work" = useless. "Fix 3 failing tests in test_banned_words.py" = actionable. Keep them short and actionable.
tags: 3–6 noun-form keywords for retrieval. Topics, not actions. Prefer "banned-words-detector" over "improving detection."
workspace: Current workspace path (absolute).
Review the session for anything worth remembering. Check activation context and recent recalls first; if a memory already exists for the subject, update it instead of creating a new one. It's fine if there's nothing worth saving — don't invent memories to look productive. If unsure, ask.