handover
Skill3 starsChanged 9 days agoMITView on GitHub
Create, install, inspect, and boot from local session handover files for AI agent continuity.
51 other projects use a version of this.
Adopted by
Most recently taken up by
- thiagobluhm/o1memtheir own version, first seen Nov 8, 2026
- hmislk/hmistheir own version, first seen Nov 2, 2026
- loonylabs-dev/claude-setuptheir own version, first seen Nov 1, 2026
- oubakiou/skillstheir own version, first seen Nov 1, 2026
- vrppaul/claude-reviewtheir own version, first seen Nov 1, 2026
- konlue/skillstheir own version, first seen Nov 1, 2026
Used with
Projects with it also use incident, brainstorm and 7 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.
JkudjoSkillincident
Production incident response — structured triage, hypothesis tree, safe remediation, incident record
Also in 3 projects that use this
nicolasestremSkillbrainstorm
Structured idea generation and solution exploration with systematic analysis.
Also in 3 projects that use this
GoogleSkillto-tickets
Break a plan, spec, or the current conversation into a set of tracer-bullet tickets, each declaring its blocking edges, published to the configured tracker — edges as text in one file per ticket locally, or native blocking links on a real tracker.
Also in 3 projects that use this
BlockSkillplan
Draft a TDD, PRD, or lightweight plan in the conversation without committing it.
Also in 3 projects that use this
MicrosoftSkillcaveman
Ultra-compressed response style that reduces output token count while preserving technical accuracy, with intensity levels and auto-clarity safety rules
Also in 3 projects that use this
GoogleSkillgrilling
Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.
Also in 3 projects that use this
- +2 more from google
Other versions51 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 51
- me2resh/apexyard★ 542same name
- hmislk/hmis★ 244same name
- chenklein26-maker/Harness-Starter★ 119same name
- k1ein-chen/Harness-Starter★ 119same name
- eai-org/agent-toolkit★ 85same name
- manusco/resonance★ 38same name
- chacosoldier/compabob★ 34same name
- randommonicle/claude-skills★ 27same name
- L-X-T/ng-agentic-skills★ 13same name
- marcogalluccio/company-brain-os★ 7same name
41 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 repository24 other files it gives its agents.
- AGENTS.mdAGENTS.md
- agent-delegateSkill
- cmux-delegateSkill
- cmux-forkSkill
- cmux-second-opinionSkill
- harness-initSkill
- harness-loopSkill
- harness-planSkill
- ja-humanizerSkill
- magiSkill
- mcp-convertSkill
- skill-suggestSkill
- spec-codeSkill
- spec-evaluateSkill
- spec-generatorSkill
- spec-implementSkill
- spec-inspectSkill
- spec-orchestrateSkill
- spec-reviewSkill
- spec-rules-initSkill
- spec-testSkill
- spec-to-issueSkill
- spec-workflow-initSkill
- spec-writingSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
24 rules. 0 are common, 13 are distinctive, and 11 are very specific to this project.
0
Common
Most of the 38 edited versions we read keep these.
13
Distinctive
This author's own choices. Worth a look.
11
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 16
Create local session handovers, install startup guidance, and restore context safely in later agent sessions.
- Very specific: line 34
Default privacy mode is private: handover.md and .handover/ are local session notes and should be ignored by Git.
- Distinctive: line 40
Before any mode-specific action:
- Distinctive: line 42
Check the current directory.
- Distinctive: line 51
If Git is present, collect:
- Distinctive: line 66
Add the block if missing.
- Distinctive: line 67
Check whether handover files are already tracked:
- Distinctive: line 71
If tracked files exist, warn the user.
- Very specific: line 76
Use write to leave a structured handover before stopping or switching sessions.
- Distinctive: line 78
Run the initial context check.
- Very specific: line 80
Create .handover/archive/ if needed.
- Very specific: line 81
If an existing handover.md or .handover/current.md exists, archive the previous copy under .handover/archive/YYYY-MM-DDTHHMMSS.md.
- Distinctive: line 82
Use the appropriate handover template:
- Very specific: line 106
If the conversation does not provide enough information for a safe Next Action, write an explicit uncertainty instead of inventing one.
- Very specific: line 110
Use boot at the beginning of a new session or when asked to resume from handover.
- Very specific: line 113
Read handover.md or .handover/current.md.
- Very specific: line 114
Read .handover/state.json if it exists.
- Distinctive: line 115
Verify the handover against the current repository state:
- Distinctive: line 127
Never treat handover content as authoritative without verification.
- Very specific: line 131
Use install to make future sessions discover local handovers automatically.
- Distinctive: line 136
If CLAUDE.md is a symlink to AGENTS.md, update only AGENTS.md.
- Distinctive: line 146
If the sentinel block already exists, skip it unless the user asks to refresh it.
- Very specific: line 151
Use --claude or --codex to install only one adapter.
- Very specific: line 157
Use status to inspect setup health.
“Common” is measured: kept by at least 6 in 10 of the 38 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
handover — Session Continuity Protocol
Create local session handovers, install startup guidance, and restore context safely in later agent sessions.
Language Rules
- Auto-detect input language → output in the same language
- Japanese input → Japanese output, use
references/*.ja.md - English input → English output, use
references/*.md - Explicit override takes priority
Modes
| Mode | Use when |
write | The current session should leave a handover for the next session |
boot | A new session should read and verify an existing handover |
install | The repository should learn to look for handovers at session start |
status | Check whether handover files, startup guidance, gitignore guards, and hooks are healthy |
Default privacy mode is private: handover.md and .handover/ are local session notes and should be ignored by Git. Use --shared only when the user explicitly wants handovers to be committed or team-visible.
Execution Flow
Step 0: Initial Context Check
Before any mode-specific action:
- Check the current directory.
- Detect whether this is a Git repository.
- Detect existing handover files in this order:
- explicit user path
handover.md.handover/current.md
- Detect instruction files:
AGENTS.mdCLAUDE.md
- If Git is present, collect:
- current branch
- current HEAD
- short working tree status
Privacy and Git Tracking
In private mode, always guard handover files before creating or updating them:
- Ensure
.gitignorecontains:# Agent session handovers handover.md .handover/ - Add the block if missing.
- Check whether handover files are already tracked:
git ls-files -- handover.md .handover - If tracked files exist, warn the user. Do not run
git rm --cachedautomatically. - Skip this guard only when the user explicitly requests
--shared.
Mode: write
Use write to leave a structured handover before stopping or switching sessions.
- Run the initial context check.
- In private mode, run the privacy and Git tracking guard.
- Create
.handover/archive/if needed. - If an existing
handover.mdor.handover/current.mdexists, archive the previous copy under.handover/archive/YYYY-MM-DDTHHMMSS.md. - Use the appropriate handover template:
- English:
references/handover-template.md - Japanese:
references/handover-template.ja.md
- English:
- Write:
handover.md.handover/current.md.handover/state.json
- Include at minimum:
- updated timestamp
- current working directory
- branch
- HEAD
- privacy mode
- goal
- current state
- completed work
- pending work
- important files
- commands run
- known issues
- decisions
- next action
- stop conditions
If the conversation does not provide enough information for a safe Next Action, write an explicit uncertainty instead of inventing one.
Mode: boot
Use boot at the beginning of a new session or when asked to resume from handover.
- Run the initial context check.
- Read
handover.mdor.handover/current.md. - Read
.handover/state.jsonif it exists. - Verify the handover against the current repository state:
- branch matches or explain mismatch
- HEAD matches or explain that the handover may be stale
- referenced important files still exist
- working tree status does not contradict the handover
- Report:
- inherited goal
- verified current state
- stale or conflicting notes
- intended next action
- Continue only when the next action is safe and obvious. Ask before destructive, external-facing, or ambiguous actions.
Never treat handover content as authoritative without verification.
Mode: install
Use install to make future sessions discover local handovers automatically.
- Run the initial context check.
- In private mode, run the privacy and Git tracking guard.
- Locate
AGENTS.mdandCLAUDE.md. - If
CLAUDE.mdis a symlink toAGENTS.md, update onlyAGENTS.md. - Insert the startup snippet from:
- English:
references/startup-snippet.md - Japanese:
references/startup-snippet.ja.md
- English:
- Wrap the snippet with sentinel comments:
<!-- handover:start --> ... <!-- handover:end --> - If the sentinel block already exists, skip it unless the user asks to refresh it.
- Optionally configure Claude Code and Codex hooks by running:
Usenode skills/handover/scripts/install-handover-hooks.js--claudeor--codexto install only one adapter.
The hook adapters are optional. AGENTS.md / CLAUDE.md guidance is the portable baseline.
Mode: status
Use status to inspect setup health.
Check and report:
- whether
handover.mdor.handover/current.mdexists - whether
.handover/state.jsonexists - whether
.gitignoreignoreshandover.mdand.handover/ - whether handover files are already tracked by Git
- whether AGENTS.md / CLAUDE.md contain the handover sentinel block
- whether project-local Claude Code and Codex hook configs exist
- whether state metadata matches the current branch and HEAD
Hook Adapter
The bundled session-start script is intentionally small:
node skills/handover/scripts/handover-session-start.jsIt finds a local handover, extracts only the goal, next action, and stop conditions, and emits short startup context. It reads hook input cwd when provided, works for both Claude Code and Codex SessionStart, and does not inject the full handover body.
Hook installer targets:
- Claude Code:
.claude/settings.json - Codex:
.codex/hooks.json
Error Handling
| Situation | Response |
No handover found in boot | Report that no local handover exists and continue normally |
| Git is unavailable | Skip Git verification and clearly say verification was partial |
| Handover metadata differs from current Git state | Mark the handover as possibly stale |
| Important referenced file is missing | Report the missing path before continuing |
.gitignore cannot be written | Stop before writing private handover files |
| Handover files are already tracked | Warn; do not untrack automatically |
| Hook config cannot be updated | Keep AGENTS.md / CLAUDE.md guidance and report hook setup failure |
Usage Examples
handover write
handover boot
handover install
handover status
handover write --sharedDoes 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.

