agentleFS
neeraj181-dotneeraj181-dot/ClaudeSkills/.claude/skills/handoff-engineer/SKILL.md

handoff-engineer

Skill0 starsChanged 3 months agoApache-2.0View on GitHub

Prepare a project so another developer or AI agent can continue work without reconstructing context — inspect architecture, current state, active work, known bugs, unfinished features, important decisions, environment requirements, and testing commands. Creates or updates HANDOFF.md with everything the next person needs to know. Based on actual project evidence, never invented. Use when the user is about to step away from a project, hands off to another developer, or wants a "cold start" document for the next session.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    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:

    Before going on vacation or switching projects.

    line 28
  2. Distinctive:

    Document problems the next person should know about:

    line 56
  3. Distinctive:

    Write the handoff document:

    line 86
  4. Distinctive:

    Every claim can be verified by reading the referenced files.

    line 145

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Handoff Engineer

Creator: Neeraj Project: ClaudeSkills

Prepare a project so the next developer or AI agent can pick up where you left off — without spending hours reconstructing context.

Core principle: the best handoff is one where the receiver never has to ask "what was going on?" A handoff document is not documentation. It is a state snapshot — what's happening right now, what's broken, what's next, and what not to touch.

This skill is not a documentation writer (documentation-engineer), not a context mapper (context-keeper), and not a decision recorder (decision-memory). It is specifically about transferring active work state to the next person.


When to activate

Activate when:

  • The user says "I need to step away" or "someone else will continue this".
  • The user says "prepare a handoff" or "write up where we are".
  • A long coding session is ending and context needs to be preserved.
  • The user wants to switch AI sessions without losing progress.
  • Before going on vacation or switching projects.

Phase 1 — Assess current state

Determine where the project stands right now:

  • Git status — what's committed, what's uncommitted, what branch are we on.
  • Recent changes — what was changed in the last N commits.
  • Build status — does it build? are there errors?
  • Test status — do tests pass? which ones fail?
  • Running state — is a dev server running? are there background processes?

Phase 2 — Identify active work

Find what was being worked on:

  • Uncommitted changes — what files are modified but not committed.
  • Untracked files — new files that haven't been committed yet.
  • Partially complete features — features that are started but not done.
  • In-progress refactors — refactoring that's partially applied.
  • Pending fixes — bugs that were being investigated.

For each, explain: what was the goal, what was done, what remains.

Phase 3 — Identify known issues

Document problems the next person should know about:

  • Bugs discovered — issues found during this session.
  • Workarounds in place — temporary fixes that need real solutions.
  • Broken features — things that used to work but don't now.
  • Test failures — tests that are failing and why.
  • Build warnings — issues that don't block but should be addressed.

Phase 4 — Document important context

Capture knowledge that only exists in the current session:

  • Why certain choices were made — reasoning that isn't in code comments.
  • What was tried and abandoned — approaches that didn't work (so they aren't retried).
  • External context — requirements, constraints, or user feedback that influenced decisions.
  • Sensitive areas — parts of the code that are fragile or dangerous to modify.

Phase 5 — Document environment requirements

Record what's needed to continue work:

  • How to start the project — exact commands.
  • How to run tests — exact commands.
  • How to build — exact commands.
  • Environment variables — what's needed (without exposing values).
  • Dependencies — anything special about the setup.
  • Ports — what ports are in use.

Phase 6 — Create HANDOFF.md

Write the handoff document:

# Project Handoff

> Generated: [date]
> Branch: [current branch]
> Status: [building / failing / testing / etc.]

## Project Overview
[One paragraph: what this project is]

## Current State
[What is the project's current status]

## What Works
[List features/functionality that is confirmed working]

## What Does Not Work
[List known broken things]

## Active Work
[What was being worked on when this handoff was created]
### What was done
### What remains
### Files involved

## Recently Changed Areas
[Files/modules modified recently]

## Important Context
[Knowledge that only exists in this session]

## Known Issues
[Bugs, workarounds, warnings]

## How to Run
### Start
[exact commands]
### Test
[exact commands]
### Build
[exact commands]

## Environment
[env vars, ports, special setup]

## What Should Be Done Next
[Prioritized list of next steps]

## Things Future Agents Must NOT Change
[Areas that are fragile, intentionally temporary, or require special care]

Phase 7 — Verify completeness

Before finishing, verify:

  • The document is based on actual evidence (not guesses).
  • Every claim can be verified by reading the referenced files.
  • The "What Should Be Done Next" is actionable and specific.
  • The "Must NOT Change" section has clear reasoning.

What this skill does NOT do

  • Does not write general documentation (that is documentation-engineer).
  • Does not map the entire codebase (that is context-keeper).
  • Does not record architectural decisions (that is decision-memory).
  • Does not fix anything — it only documents the current state.

This skill answers one question: "What does the next person need to know to pick this up immediately?"


Final output

# Handoff Summary
# Project State
# Active Work
# Known Issues
# Environment
# Next Steps
# Critical Warnings

Guardrails: base everything on actual project evidence; never invent project history; never expose secrets; document what is broken, not just what works; keep the document concise and actionable; verify every claim against actual files; and never claim something is "done" without checking.

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.