Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsBlogPro
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges
  • Chrome Extension
  • Skill Manager

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Compact Handoff

ASecurity

Generate exactly two ready-to-paste prompts from the current conversation, labeled COMPACT_PROMPT and COMPACT_POST_PROMPT: a `/compact` command that carries the working state forward, and a post-compaction prompt that directs the next unresolved task. Prints text only; executes neither prompt. TRIGGER when: /compact-handoff is invoked, or the operator asks for a compact prompt plus a resume prompt, a compaction carry-forward, or a handoff before compacting.

40 stars
0 votes
0 copies
0 views
Added 9/29/2026
ai-agentsgo

Security Analysis

A100/100

Scanned 9/29/2026

$npx -y skills add mifunedev/agro --skill compact-handoff --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Compact Handoff?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Compact Handoff
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/mifunedev-compact-handoff/badge)](https://www.skillsdirectory.com/skills/mifunedev-compact-handoff)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
Files
SKILL.md
---
name: compact-handoff
description: |
  Generate exactly two ready-to-paste prompts from the current conversation,
  labeled COMPACT_PROMPT and COMPACT_POST_PROMPT: a `/compact` command that
  carries the working state forward, and a post-compaction prompt that directs
  the next unresolved task. Prints text only; executes neither prompt.
  TRIGGER when: /compact-handoff is invoked, or the operator asks for a compact
  prompt plus a resume prompt, a compaction carry-forward, or a handoff before
  compacting.
argument-hint: "[focus]"
disable-model-invocation: true
---

# Compact Handoff

Arguments received: `$ARGUMENTS`

Produce two prompts. Execute neither. Change no files, run no tools that
mutate state, and dispatch no workers.

## 1. Collect the state

Read the conversation from the start. Weight later messages over earlier ones.
If `$ARGUMENTS` is not empty, treat it as the focus for the next task and for
what to keep. The focus adds no requirements.

Extract each item below. Keep only what a fresh context needs to continue.

- **Objective:** the overall goal, in the operator's terms.
- **Latest instructions:** the most recent operator directions. A later
  instruction replaces an earlier one on the same point.
- **Decisions:** choices the operator made or approved.
- **Constraints:** hard rules, boundaries, and exclusions in force.
- **Completed work:** what is done, with the evidence that proves it.
- **Essential references:** file paths, branches, PRs, issues, commands, URLs,
  and identifiers that the next step needs. Copy them exactly.
- **Unresolved next steps:** open work, in order, and any open question
  blocked on the operator.

Classify every candidate item before you keep it:

| Status | Treatment |
|---|---|
| Approved by the operator | Keep as a decision. |
| Proposed but not approved | Keep only as an open proposal, labeled as such. |
| Superseded, abandoned, or answered | Drop. |
| Tool output, file dumps, exploration, retries | Drop. Keep only the conclusion and its reference. |

Do not invent a requirement, criterion, path, or decision. If a value that
the next task needs is missing, write `<unknown: what is missing>`.

## 2. Select the next task

Pick the first unresolved step that waits on no operator answer. If every
remaining step waits on the operator, make the next task "ask the operator" and name the
blocking question. If no work remains, make the next task "confirm completion
with the operator".

## 3. Write the prompts

**`COMPACT_PROMPT` — the `/compact` command.** Write one command that the
operator pastes as is. Start with `/compact ` and follow it on the same line
with the compaction instruction. The instruction tells the summarizer what to
keep to deliver the best result on the next task:

- the seven items from step 1, each with its concrete values;
- open proposals, labeled as unapproved;
- the references that the next task needs, copied exactly.

The instruction also tells the summarizer to drop stale, superseded, and
redundant context, raw tool output, and file dumps.

**`COMPACT_POST_PROMPT` — the post-compaction prompt.** Write the message that
the operator sends after the compaction. The message directs the task from
step 2 and states:

- **Objective:** what the task achieves.
- **Scope:** what is in and out.
- **Deliverable:** the concrete output.
- **Completion criteria:** binary checks that prove the task is done.

Point at files for re-anchoring instead of restating their contents. Take the
criteria from the conversation. If the conversation does not define one, use
`<unknown: criterion>`.

## 4. Output

Print exactly this shape and nothing else:

````text
COMPACT_PROMPT:
```
/compact <compaction instruction>
```
COMPACT_POST_PROMPT:
```
<post-compaction prompt>
```
````

Use a plain fence with no language tag for each prompt. If a prompt contains
a triple backtick, fence that prompt with four backticks. Print no preamble,
commentary, or closing text.

## Examples

- `/compact-handoff` — prompts for the current conversation.
- `/compact-handoff finish the CI fix` — same, with the next task focused on
  the CI fix.

Attribution

mifunedevmifunedev
View sourceSee grades on GitHubMore from mifunedev →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Terse caveman voice: answer first, fluff gone, every technical fact kept. Use for /caveman, "caveman mode", "talk like caveman", "be brief", "less tokens". Stays on until "stop caveman" or "normal mode".

1100021 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

698621 votes

Writing Skills

Create and manage Claude Code skills in HASH repository following Anthropic best practices. Use when creating new skills, modifying skill-rules.json, understanding trigger patterns, working with hooks, debugging skill activation, or implementing progressive disclosure. Covers skill structure, YAML frontmatter, trigger types (keywords, intent patterns), UserPromptSubmit hook, and the 500-line rule. Includes validation and debugging with SKILL_DEBUG. Examples include rust-error-stack, cargo-dep...

3931 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3421 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Amp, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Grok Build, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

741 votes
View all in ai-agents →