wrap-up

Distills session work into wiki journals, dashboards, and staged knowledge entries.

Updated May 14, 2026
One-click install
npx skills add https://github.com/dadlabs-io/llm-wiki-bootstrap --skill wrap-up-dadlabs-io
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: wrap-up
Source: https://github.com/dadlabs-io/llm-wiki-bootstrap/tree/main/bootstrap/skills/wrap-up
Command: npx skills add https://github.com/dadlabs-io/llm-wiki-bootstrap --skill wrap-up-dadlabs-io

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? End-of-session context is lost when a conversation ends: decisions, bug fixes, and progress live only in the transcript. This Skill crystallizes the current session into a project wiki so the next session can resume in seconds and durable knowledge is preserved. ## Core Features & Use Cases - Session journal upsert: Maintains one append-only journal per session at wiki/sessions/<persona>/<YYYY-MM>/, so repeat wrap-ups build a single running record. - Working-memory dashboards: Refreshes handoff.md, task.md, and active-context.md, the resume pointers a cold-start session reads first. - Durable knowledge staging: Extracts components, decisions, architecture, patterns, and troubleshooting entries into _inbox/proposed/ with metadata sidecars, then offers inline promotion to wiki/project/<category>/. - Optional commit and push: With --auto-commit or --auto-push (or the wrap_up_commit notebook setting), it commits only the session's changed paths and pushes branches that already track a remote. - Use Case: After a two-hour debugging session, run /wrap-up to log what happened, update the resume dashboards, file the root-cause fix as a troubleshooting entry, and commit everything before closing the window. ## Quick Start Ask the AI to wrap up this session and document what we did in the project wiki.

Frequently Asked Questions about wrap-up

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I document what was done in an AI coding session?▼

Run /wrap-up at the end of the session. It writes a per-session journal entry, refreshes the handoff and task dashboards, and stages distilled entries for decisions, components, patterns, and bugs into the project wiki.

What is the difference between wrap-up and wiki-update?▼

wrap-up captures your own session work into project/ categories like decisions and troubleshooting. wiki-update ingests external sources such as URLs, papers, and videos into research/. A pure research session should use wiki-update instead.

Can wrap-up commit and push my changes automatically?▼

Yes. Use /wrap-up --auto-commit to commit the session's changed paths, or --auto-push to also push branches that already track a remote. The per-notebook wrap_up_commit setting (none, commit, or push) can make this the default.

Does wrap-up work on a brand-new repository with no commits?▼

Yes. When git log fails on a repo with no commits, it falls back to git status and the conversation itself as the primary source. Everything is treated as new, which is the normal first-wrap-up state.

When should I not run wrap-up?▼

Skip it for trivial sessions like typo fixes or dependency bumps, and for pure external-source ingestion where wiki-update fits better. For sessions with no settled output, it reports that nothing merits a wiki entry rather than filing thin ones.