agentleFS
wickedhardflipwickedhardflip/claude-code-tips/skills/wrap/SKILL.md

wrap

Skill1 starsChanged 38 days agoMITView on GitHub

Use when the user types /wrap or asks to wrap up, save progress, or checkpoint the session before stopping, rebooting, or switching tasks.

  • Reads credentials

46 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use start, update and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • AnthropicSkill

    start

    Set up your bio-research environment and explore available tools. Use when first getting oriented with the plugin, checking which literature, drug-discovery, or visualization MCP servers are connected, or surveying available analysis skills before starting a new project.

    Also in 5 projects that use this

  • AnthropicSkill

    update

    Sync tasks and refresh memory from your current activity. Use when pulling new assignments from your project tracker into TASKS.md, triaging stale or overdue tasks, filling memory gaps for unknown people or projects, or running a comprehensive scan to catch todos buried in chat and email.

    Also in 3 projects that use this

  • LangChainSkill

    remember

    Review the current conversation and capture valuable knowledge — best practices, coding conventions, architecture decisions, workflows, and user feedback — into persistent memory (AGENTS.md) or reusable skills. Use when the user says: (1) remember this, (2) save what we learned, (3) update memory, (4) capture learnings.

    Also in 3 projects that use this

  • termdockSkill

    recall

    Retrieves cross-session memories about past decisions, lessons, and patterns. Use when user asks about prior work, past decisions, or mentions a feature. Triggers include check memory, what did we decide, how did we solve.

    Also in 3 projects that use this

  • Microsoft AzureSkill

    build

    Build the TypeSpec Rust emitter from source. Run this after cloning the repo, after pulling changes, or when TypeScript source files have been modified.

    Also in 3 projects that use this

  • SentrySkill

    review

    Fixture review skill.

    Also in 7 projects that use this

  • +1 more from anthropics

The file

We split this skill into the rules it gives an agent

9 rules. 0 are common, 5 are distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 40 edited versions we read keep these.

  • 5

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use the memory directory and format from the system prompt.

    line 24
  2. Distinctive:

    Update the existing file for a project or topic.

    line 25
  3. Very specific:

    Keep the MEMORY.md index in sync: one line per file, no duplicates.

    line 28
  4. Very specific:

    If a project had multi-step work but no resume doc, create TASKS.md at its root.

    line 38
  5. Very specific:

    If the workspace has logs/journal.md, append one dated entry (## YYYY-MM-DD: <topic>): 3–6 bullets covering decisions, progress, blockers, …

    line 42
  6. Very specific:

    Never stage secrets: .env, keys, tokens, credential or config files with passwords.

    line 49
  7. Distinctive:

    Commit on the current branch with a clear message and the session's attribution lines.

    line 50
  8. Distinctive:

    Never push: Never force, amend, or rebase.

    line 51
  9. Distinctive:

    Stop background tasks that shouldn't outlive the session (servers, watchers), unless the user wants them running.

    line 58

“Common” is measured: kept by at least 6 in 10 of the 40 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

Wrap

Checkpoint the session so a fresh session can resume with no scrollback. Capture state in the places that already exist, commit local repo work, and report.

Work through every step. A step with nothing to do gets one line in the report ("Journal: n/a, no journal in this workspace").

1. Inventory (no writes yet)

Note, from this conversation:

  • Projects touched: each directory where files were created or edited.
  • Decisions: what was chosen, and why it matters later.
  • State: what's done, what's in progress, what's blocked (and on what).
  • Next step: the single concrete action to resume with.
  • Loose ends: background tasks/servers/downloads still running; actions waiting on the user (reboot, login, approval).

2. Memory

Use the memory directory and format from the system prompt.

  • Update the existing file for a project or topic. Create a new one only for a genuinely new fact, preference, or project.
  • Save user preferences and corrections as feedback memories, with Why and How to apply.
  • Store absolute dates, never "today" or "tomorrow".
  • Keep the MEMORY.md index in sync: one line per file, no duplicates.
  • Skip anything the repo, docs, or git history already record.

3. Project resume docs

For each project touched, update the resume docs it already has: NEXT.md, TASKS.md, CONTEXT.md, plan files with checkboxes, and status sections in a project CLAUDE.md.

  • Tick finished items and add newly found ones.
  • Move the ← START HERE marker (or equivalent) to the next step.
  • Record blockers and exactly how to clear them.

If a project had multi-step work but no resume doc, create TASKS.md at its root. Otherwise create nothing new.

4. Journal

If the workspace has logs/journal.md, append one dated entry (## YYYY-MM-DD: <topic>): 3–6 bullets covering decisions, progress, blockers, and the resume pointer.

5. Git (every repo touched)

For each touched directory inside a git repo:

  1. git status. Read .gitignore if unsure what belongs.
  2. Stage only files from this session's work, by explicit path. Never git add -A/..
  3. Never stage secrets: .env, keys, tokens, credential or config files with passwords. If one shows as modified, leave it out and flag it in the report.
  4. Commit on the current branch with a clear message and the session's attribution lines.
  5. Never push. Never force, amend, or rebase.
  6. Report unrelated uncommitted changes you left alone.

A repo that is public (per memory or CLAUDE.md) gets an extra secrets check of the staged diff before committing.

6. Loose ends

Stop background tasks that shouldn't outlive the session (servers, watchers), unless the user wants them running. Leave user-started processes alone. List everything still running or pending.

7. Report (this exact shape)

**Wrapped.**
- Memory: <files created/updated, or n/a>
- Docs: <files updated, or n/a>
- Journal: <entry added, or n/a>
- Git: <repo: commit hash + message | nothing to commit | not a repo>
- Still running / needs you: <items, or none>

**To resume:** say "<short phrase>". Starts at <file / next step>.

Common Mistakes

MistakeFix
Creating a new notes file when a NEXT.md/TASKS.md existsUpdate the existing doc
git add . sweeps in unrelated or secret filesStage explicit paths from this session only
Pushing "since it's committed anyway"Never push; the user pushes
Memory duplicates an existing fileSearch the memory dir first; update in place
Relative dates ("tomorrow") in memory/docsConvert to YYYY-MM-DD
Skipping a step silentlyEvery step appears in the report, even as n/a

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.