agentleFS
nfeuernfeuer/project-foundation-kit/.claude/skills/session-handoff/SKILL.md

session-handoff

Skill0 starsChanged 4 months agoMITView on GitHub

Write a tight durable state summary so a fresh session or another agent can resume the current task without re-deriving context

149 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use ascii-diagram, research-fleet 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.
  • SethGammonSkill

    ascii-diagram

    Generate perfectly aligned ASCII diagrams — architecture, flow, sequence, box-and-arrow. Uses a programmatic character-grid approach so alignment is guaranteed by math, not token prediction. Includes post-render verification.

    Also in 4 projects that use this

  • SethGammonSkill

    research-fleet

    Parallel research using Fleet wave mechanics. Spawns multiple scout agents, each investigating a different angle of the same question. Findings are compressed between waves. Produces a unified research brief from multiple independent perspectives.

    Also in 3 projects that use this

  • vicoa-aiSkill

    live-preview

    Start a local app in the current user project, expose it through a tunnel, and return preview details in structured Markdown. Use when the user asks for a preview URL, public URL, shareable localhost, mobile/web preview, or tunnel setup.

    Also in 4 projects that use this

  • botlearn-aiSkill

    doc-gen

    Also in 4 projects that use this

  • flyingobsidianSkill

    terminology

    Read a codebase, find terminology, create/update definitions into a markdown document.

    Also in 3 projects that use this

  • Dizzy-TenSkill

    bug-diagnosis

    Diagnose bugs and product issues without modifying code. Use when the user explicitly invokes $bug-diagnosis or asks for diagnosis, root-cause analysis, issue classification, or fix strategy for code bugs, failing tests, configuration/dependency problems, deployment issues, performance regressions, data anomalies, or frontend/user-experience problems described through screenshots, annotations, logs, errors, or natural-language reports. Prioritize analysis and recommendations only; do not edit files.

    Also in 3 projects that use this

  • +1 more from SethGammon

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 127 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run this when a long task is being interrupted — end of session, context window pressure, handing off to a parallel agent, or stepping away …

    line 15
  2. Distinctive:

    Write the completed note to the handoff slot:

    line 74
  3. Distinctive:

    Never include secrets, tokens, or full file contents in the handoff note.

    line 92
  4. Distinctive:

    If the task is complete, do not write a handoff note — close out the task and delete any existing handoff file instead.

    line 93

“Common” is measured: kept by at least 6 in 10 of the 127 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

Session Handoff

Profile-driven. The protected trunk branch is trunk_branch and worktree isolation lives under worktree_dir in .claude/kit.yaml; reference both in the handoff note so the next agent targets the right branch and worktree path.

Run this when a long task is being interrupted — end of session, context window pressure, handing off to a parallel agent, or stepping away mid-slice. The goal is a baton, not a report: the minimum state a fresh agent needs to pick up and run, without re-reading every file you already read. Keep it under one screen. If it's longer than that, you've written a report; trim it.

Write the handoff note to the project's scratchpad or a known slot in the repo (docs/handoff.md or .claude/handoff.md). The next session's first instruction should be: "read .claude/handoff.md before doing anything." Delete or overwrite the file when the task completes.

Workflow

1. Capture the task identity

One sentence: what problem are we solving, and why does it matter now? Include the slice name or issue number if applicable.

2. List what is done

Bullet each completed sub-task. Include commit SHAs or PR numbers so the next agent can verify without re-running anything.

git log --oneline -10   # grab the relevant SHAs
git status --short      # confirm working tree state

3. Record what is in progress

Which sub-task is currently half-done? What file is it in? What is the next line to write or the next test to pass?

4. Enumerate touched files and the active branch

git diff main...HEAD --name-only   # files changed from trunk_branch (kit.yaml)
git branch --show-current           # active branch
git worktree list                   # worktrees live under worktree_dir (kit.yaml)

5. Summarize key decisions and their rationale

Two to five bullets, each following the pattern: <decision> — because <reason>. Only include decisions that aren't obvious from reading the code; things a fresh agent might reverse by mistake.

6. State the next concrete step

Be specific enough that the next agent can act without any inference: file path, function name, what to add or change.

7. List blockers

Anything the next agent cannot proceed past without human input or an external dependency resolving. If there are none, say so explicitly.

8. Write the file

Write the completed note to the handoff slot:

# project-local slot (preferred)
cat > .claude/handoff.md << 'EOF'
<handoff note>
EOF

# or scratchpad if the working tree is sensitive
cat > /tmp/claude-handoff.md << 'EOF'
<handoff note>
EOF

Guardrails

  • The handoff note is machine-readable state, not a progress narrative. Omit context the next agent can derive from reading the files you list.
  • Never include secrets, tokens, or full file contents in the handoff note.
  • If the task is complete, do not write a handoff note — close out the task and delete any existing handoff file instead.

Output

## Handoff — <task name / slice ID> — <timestamp>

**Task:** <one sentence: what and why>
**Branch:** <branch-name> (trunk: `trunk_branch`) [worktree: <path under worktree_dir if applicable>]

### Done
- <sub-task> — <commit SHA or PR #>
- <sub-task> — <commit SHA or PR #>

### In Progress
- <sub-task> — half-done in `<file:line>`, need to <specific next action>

### Key Decisions
- <decision> — because <reason>
- <decision> — because <reason>

### Next Step
`<file path>` → `<function/class>`: <exactly what to write or change>

### Blockers
- <blocker> — waiting on <who/what>
- none

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.