load-handover
Skill16 starsChanged 57 days agoMITView on GitHub
Load the most recent handover note from the previous session into context.
In this repository23 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- adviseSkill
- brainstormSkill
- code-navigationSkill
- commitSkill
- create-agents-mdSkill
- create-prSkill
- design-buildSkill
- design-fixSkill
- designSkill
- handoverSkill
- hypothesis-debuggingSkill
- local-filesSkill
- planSkill
- report-checkSkill
- reviewSkill
- senior-reviewSkill
- shipSkill
- tddSkill
- verificationSkill
- visual-companionSkill
- workSkill
The file
We split this skill into the rules it gives an agent
9 rules. About 0 are common, 5 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
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
- Very specific: line 18
If the listing shows an error (e.g., "No such file or directory"), is empty, or contains no .md files:
- Very specific: line 24
If there are one or more .md files:
- Very specific: line 26
Read it using the Read tool at path .claude/handovers/{filename}.
- Very specific: line 28
If more than one .md file exists, append: "{count} older handover(s) also available."
- Distinctive: line 34
After reading the handover file, present this summary:
- Distinctive: line 40
If the handover has no "Next Steps" section or it says "N/A", set Top Priority to "No next steps recorded" and ask the user what they'd …
- Distinctive: line 45
Ask the user which one they'd like to work on.
- Distinctive: line 53
Don't: read older handover files unless the user explicitly requests it.
- Distinctive: line 60
If the file content looks malformed (no section headings, empty body), warn: "This handover may be incomplete — proceeding with available …
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Previous Session Handover
Instructions
First, use the Glob tool to list .claude/handovers/*.md. Sort results by filename descending (newest first, since filenames are timestamps).
Using the Glob results, determine which branch applies:
Branch A — No Files
If the listing shows an error (e.g., "No such file or directory"), is empty, or contains no .md files:
No handover files found for this project. This is a fresh session — no previous context to load. Stop here.
Branch B — Files Exist
If there are one or more .md files:
- The first
.mdfile in the listing is the most recent (sorted newest-first). Read it using the Read tool at path.claude/handovers/{filename}. - Present it using the output template below.
- If more than one
.mdfile exists, append: "{count} older handover(s) also available."
Output Template
After reading the handover file, present this summary:
Session loaded:
{filename}Date: {date extracted from timestamp in filename} Top Priority: {first item from Next Steps section}
If the handover has no "Next Steps" section or it says "N/A", set Top Priority to "No next steps recorded" and ask the user what they'd like to focus on.
Then:
- Confirm you understand the current state of the project.
- List all Next Steps with their priority order and a short explanation of each.
- Ask the user which one they'd like to work on.
Anti-Patterns
- Don't dump the raw handover contents without a summary header — always lead with the output template.
- Don't auto-start the first Next Step without asking — always present the list and let the user choose.
- Don't read older handover files unless the user explicitly requests it.
Verification
- Confirm the file read succeeded (non-empty content returned).
- If the file content looks malformed (no section headings, empty body), warn: "This handover may be incomplete — proceeding with available context."
- Confirm the output template was fully populated — no raw
{placeholder}text remains in your response.
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.

