wrap
Skill0 starsChanged 9 months agoMITView on GitHub
Coordinates the end of a session by chaining sync-docs, an instruction-doc quality pass (CLAUDE.md + AGENTS.md), and handoff into one /wrap command; optionally adds a handoff-fresh bundle. Captures the verbatim last exchange first so nothing is lost. Do NOT use for a bare handoff (use /handoff) or to resume/ingest a prior session (use /pickup).
46 other projects use a version of this.
Adopted by
Most recently taken up by
- odakin/claude-configtheir own version, first seen Nov 15, 2026
- fjwood69/moritheir own version, first seen Nov 10, 2026
- rhasselle-oddball/claude-skillstheir own version, first seen Nov 8, 2026
- jolleyDesign/claude-skillstheir own version, first seen Nov 8, 2026
- egregore-labs/egregoretheir own version, first seen Nov 6, 2026
- mag123c/toktracktheir own version, first seen Nov 3, 2026
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.
AnthropicSkillstart
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
AnthropicSkillupdate
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
LangChainSkillremember
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
termdockSkillrecall
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 AzureSkillbuild
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
SentrySkillreview
Fixture review skill.
Also in 7 projects that use this
- +1 more from anthropics
Other versions46 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 46
- UniClipboard/UniClipboard★ 2ksame name
- team-attention/stanford-cs146s-kr★ 294same name
- egregore-labs/egregore★ 293same name
- mag123c/toktrack★ 196same name
- neurofoo/agent-skills★ 111same name
- lsnackerman/raw-clauding-skills★ 45same name
- fjwood69/mori★ 32same name
- YahyaLendani/opus-tuner-claude-skill★ 14same name
- jackjin1997/ClawForge★ 12same name
- ucsandman/claude-skills★ 5same name
36 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository17 other files it gives its agents.
- AGENTS.mdAGENTS.md
- auditing-skillsSkill
- creating-skillsSkill
- forge-builderSkill
- forge-claude-teamsSkill
- forge-codex-multiagentSkill
- forge-orchestratorSkill
- forge-researchSkill
- handoff-freshSkill
- handoffSkill
- managing-doc-manifestSkill
- okfSkill
- pickupSkill
- studySkill
- syncing-docsSkill
- testing-skillsSkill
- updating-skillsSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
13 rules. 0 are common, 9 are distinctive, and 4 are very specific to this project.
0
Common
Most of the 40 edited versions we read keep these.
9
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
- Distinctive: line 33
Do not run via automatic hooks unless user explicitly configures and documents that behavior
- Very specific: line 48
Before running sync (which can be long enough for a context compaction to occur), capture the last user prompt and last assistant response …
- Distinctive: line 56
No args: Run all three steps in sequence
- Distinctive: line 66
Check cross-file consistency across all state files
- Distinctive: line 76
Keep shared execution context aligned with CLAUDE.md
- Very specific: line 95
Only when user explicitly passes --with-fresh:
- Very specific: line 111
If a .handoff-fresh/ bundle already exists but you wrote only the canonical HANDOFF.md this run, say so — the now-older bundle is …
- Distinctive: line 115
Never leave both a fresh single-file handoff and an older bundle with no note of which is authoritative.
- Distinctive: line 137
Update this skill when:
- Distinctive: line 138
New end-of-session skill added: Add to the chain
- Distinctive: line 139
Ordering issue: If a step should run before/after another, adjust sequence
- Very specific: line 140
Fresh-agent flow changes: Update --with-fresh step and output contract
- Distinctive: line 141
Instruction-doc drift: If CLAUDE.md and AGENTS.md drift repeatedly, tighten parity pass in Step 2
“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.
Files
Wrap
Chains end-of-session skills into one command: sync docs, audit instruction docs quality (CLAUDE.md + AGENTS.md), and hand off.
/wrap closes a session. Its consumer counterpart is /pickup, which
opens the next one — it reads whatever /wrap (via /handoff /
/handoff-fresh) wrote. Together they form a producer→consumer contract pair.
Command Contract (Explicit, Manual)
- Primary entry point:
/wrap - Execute only when user asks explicitly (
/wrapor direct equivalent wording) - Do not run via automatic hooks unless user explicitly configures and documents that behavior
- Optional fresh bundle generation must be explicit (
--with-fresh)
Workflow
- [ ] 0. Capture the verbatim last exchange FIRST (before sync/compaction can lose it)
- [ ] 1. Run syncing-docs (full mode — drift fix + session learnings + cross-file checks)
- [ ] 2. Run instruction-doc quality pass (CLAUDE.md improver + AGENTS parity mirror)
- [ ] 3. Run handoff (commit, update docs, generate HANDOFF.md)
- [ ] 4. Run handoff-fresh (optional, only if --with-fresh)Step 0: Capture the Verbatim Last Exchange First
Before running sync (which can be long enough for a context compaction to occur),
capture the last user prompt and last assistant response verbatim into
working notes, per the /handoff Step 0 contract. This guarantees the verbatim
resume anchor survives into the handoff regardless of what happens during the
later steps. Capture is manual — /wrap installs no auto-hooks.
Arguments
- No args: Run all three steps in sequence
--no-handoff: Skip step 3 (useful mid-session when you want to sync but keep working)--quick: Pass--quickto syncing-docs (drift fix only, skip learnings/cross-file)--with-fresh: After step 3, run/handoff-fresh --no-syncto generate fork-safe onboarding bundle
Step 1: Sync Docs
Invoke the syncing-docs skill (full mode unless --quick):
- Fix drift between code and owned docs
- Capture session learnings into CLAUDE.md + AGENTS.md
- Check cross-file consistency across all state files
Step 2: Audit Instruction-Doc Quality
Invoke the claude-md-improver skill for CLAUDE.md:
- Score CLAUDE.md against quality rubric
- Fill gaps (missing commands, architecture, gotchas)
- Apply improvements with user approval
Then mirror shared-context quality-critical updates into AGENTS.md:
- Keep shared execution context aligned with CLAUDE.md
- Preserve platform-specific sections where needed
- Flag any unresolved CLAUDE/AGENTS contradiction in final summary
Step 3: Handoff
Invoke the handoff skill (v3.0.0):
- Commit current work
- Generate/update HANDOFF.md with session snapshot, including the mandatory
## Last Exchange (Verbatim)and machine-runnable## Verify Blocksections (and the line-1 schema marker) — these are what make the handoff consumable by/pickup - Ready for /clear or session end
To resume in the next session, the user runs /pickup (or "read HANDOFF.md",
which routes to /pickup).
Step 4: Fresh Bundle (Optional)
Only when user explicitly passes --with-fresh:
- Invoke handoff-fresh with
--no-sync - Generate fresh-agent onboarding files in
.handoff-fresh/current/ - Report bundle output paths in final summary
Choosing the artifact shape (avoid silent inconsistency)
The output shape must be a deliberate choice, not an accident of which flag was
typed. Default /wrap writes the single canonical HANDOFF.md — correct for
same-repo continuity. Recommend --with-fresh (the multi-file bundle) when the next
agent is fork/clone/fresh-onboarding (new worktree, handed to a teammate, or a
cold start with no prior context).
- Always state the shape in the final summary: which artifact was written
(
HANDOFF.mdvs.handoff-fresh/bundle) and that/pickupconsumes it. - If a
.handoff-fresh/bundle already exists but you wrote only the canonicalHANDOFF.mdthis run, say so — the now-older bundle is historical (handoff's precedence header +/pickup's freshness check keep the next resume from following the stale bundle). - Never leave both a fresh single-file handoff and an older bundle with no note of which is authoritative.
Example
/wrapRuns sync-docs → claude-md-improver → handoff sequentially. Each step shows its own output. Final output:
=== Wrap Complete ===
Sync: 4 drift fixes (CLAUDE.md + AGENTS.md), 1 learning captured, 2 cross-file issues flagged
Quality: CLAUDE.md scored B (78/100), 2 improvements applied, AGENTS.md parity mirrored
Handoff: Committed (abc1234), HANDOFF.md updated
Artifact shape: single canonical HANDOFF.md (no bundle) — /pickup will consume it
Fresh: (omitted unless --with-fresh)Self-Evolution
Update this skill when:
- New end-of-session skill added: Add to the chain
- Ordering issue: If a step should run before/after another, adjust sequence
- Fresh-agent flow changes: Update
--with-freshstep and output contract - Instruction-doc drift: If CLAUDE.md and AGENTS.md drift repeatedly, tighten parity pass in Step 2
Applied Learnings:
- v1.5.0: Artifact-shape discipline (gap NEW-B from the A/B eval). Made the
single-file-vs-bundle choice deliberate and self-described: guidance on when to add
--with-fresh, the final summary always states which artifact was written and that/pickupconsumes it, and a coexisting older bundle is explicitly marked historical (backed by handoff's new precedence header). Fixes the silent "shape depends on which flag was typed" inconsistency. - v1.4.0: Added Step 0 (capture the verbatim last exchange first, before sync, so a compaction during later steps cannot lose it) and documented the wrap↔pickup producer→consumer contract pair. Step 3 now relies on handoff v3.0.0's mandatory Last Exchange + Verify Block sections that make the handoff consumable by
/pickup. - v1.3.0: Step 2 expanded from a CLAUDE-only quality pass to an instruction-doc quality flow (CLAUDE.md improvements + AGENTS.md parity mirror).
- v1.2.0: Updated
--with-freshto target the foldered handoff-fresh bundle path.handoff-fresh/current/. - v1.0.0: Initial end-of-session coordinator chaining sync-docs → claude-md-improver → handoff.
Current version: 1.5.0. See CHANGELOG.md for history.
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
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.

