agentsclimarketplace

Handoff

Skill arendon1/agent-skills/process/handoff

A curated collection of autonomous AI agent skills powered by Vercel's `skills` npm cli package. Specialized tools for productivity, document management, and custom LLM workflows.

Install
npx -y skills add arendon1/agent-skills --skill handoff

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Compact the current session into a handoff document so another agent (or a fresh session) can continue the work. Captures goal, state, decisions, next steps, and pointers to artifacts. Does not duplicate content already in artifacts. Use when ending a session mid-plan, or when handing work to another agent, or when the user says "handoff", "compact this", "I'm leaving", "pass this to another agent", "save state for later", "wrap up".

SKILL.md

4.1 KB, 932 tokens by cl100k_base, as published. Nobody here has run it

handoff — compact the session

Compact the current conversation into a handoff document so a fresh agent can continue the work. Captures goal, state, decisions, next steps, and pointers to artifacts — without duplicating content already captured in plan artifacts, commits, diffs, or ADRs.

OWNERSHIP

Owns: the handoff document (saved to the OS temp dir, NOT the workspace). Reads: the active plan's artifacts (PRD.md, ARD.md, SPEC.md, PLAN.md, LESSONS.md, RESEARCH.md), CONTEXT.md, git state. MUST NOT write plan artifacts. References them by path, never duplicates them.

WHEN (user-invoked)

  • Ending a session mid-plan.
  • Handing work to another agent.
  • User says "handoff", "compact this", "I'm leaving", "pass this to another agent", "save state for later", "wrap up".
  • Before a compaction resets context (if invoked manually rather than by the adapter).

STEP 0 — CAPTURE THE ARGUMENT

If the user passed an argument, treat it as a description of what the next session will focus on and tailor the handoff to it. Otherwise, capture the whole current state.

STEP 1 — GATHER STATE

  1. Resolve the active plan per §6. Note its path.
  2. Read the active plan's artifacts. Note their paths + a one-line summary of each (do NOT duplicate their content).
  3. Read CONTEXT.md. Note its path.
  4. Check git state: current branch, uncommitted changes, recent commits.
  5. Note any ? items, in-progress tasks (~), and blocking unknowns.

STEP 2 — WRITE THE HANDOFF

Save to the OS temp directory (not the workspace — it is a scratch doc, not a versioned artifact). Use caveman-light (compressed but readable).

Template:

# HANDOFF — <objective>

> For a fresh agent picking up <objective>. Read this, then open the plan artifacts.

## goal
<one line: what the work is>

## current state
- active plan: docs/plans/<date>-<type>-<slug>/
- branch: <branch>, <clean|uncommitted: N files>
- artifacts: <list paths + one-line summary each, NOT content>

## decisions made
- <decision 1> (in ARD.md §D / ADR 000N)
- <decision 2>

## next steps
1. <next step 1 — e.g. "run build on §T.4">
2. <next step 2>
3. <open question parked as ? — needs research>

## pointers
- plan artifacts: docs/plans/<slug>/
- CONTEXT.md: ./CONTEXT.md
- recent commits: <commit hashes + subjects>

## suggested skills
- <skill 1> — <why>
- <skill 2> — <why>

## redacted
- (any sensitive info redacted — API keys, passwords, PII)

SOURCE DISCIPLINE

  • DO NOT duplicate content already captured in artifacts (PRDs, plans, ADRs, issues, commits, diffs). Reference them by path or URL instead.
  • Redact any sensitive information: API keys, passwords, personally identifiable information. If you must reference that a secret exists, say "secret configured in env var FOO_KEY" without the value.

SUGGESTED SKILLS

Include a "suggested skills" section that names the skills the next agent should invoke, with a one-line reason each. Examples:

## suggested skills
- bootstrap — re-enable auto-triggering + artifact-check habit at session start
- build — execute §T.4 next (status ~, blocked by §T.3 which is [x])
- verify — gate §T.4 before flipping to [x]
- lessons — if §T.4's test fails, trace root cause

BOUNDARIES

  • MUST save to the OS temp dir, NOT the workspace.
  • MUST NOT duplicate content already in artifacts. Reference by path.
  • MUST redact sensitive information.
  • MUST include a "suggested skills" section.
  • MUST use canonical terms from CONTEXT.md.
  • MUST NOT write plan artifacts.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 2 of the 12 instructions most agent orchestration skills give in 932 tokens

Counted across 742 of the 995 authors here whose files we hold, read 2026-08-07

  • Reference existing artifacts by path or URLin 53 of 742, across 25 files
  • Run the full test suite after integrating changesin 51 of 742, across 19 files
  • Dispatch one agent per independent problem domainin 50 of 742, across 17 files
  • Verify fixes do not conflictin 45 of 742, across 13 files
  • Include a suggested skills section in the documenthere, and in 45 of 742, across 17 files
  • Redact sensitive informationhere, and in 41 of 742, across 11 files
  • Save to the temporary directory of the operating systemin 39 of 742, across 10 files
  • Tailor the document to user-provided focus argumentsin 39 of 742, across 9 files
  • Spot check agent changes for systematic errorsin 34 of 742, across 7 files
  • Write a handoff document summarising the current conversationin 31 of 742, across 6 files
  • Assign each agent a specific scopein 23 of 742, across 8 files
  • Provide specific scope and clear goalin 23 of 742, across 5 files

Said here and by no other author read

  • read active plan artifacts and note paths
  • note in-progress tasks and blocking unknowns
  • use canonical terms from context configuration
  • tailor the handoff to any user-provided focus argument

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.