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

Cavecrew

ASecurity

决定何时把工作委派给 caveman 风格 subagent 的判断指南。告诉主线程 什么时候该派 `cavecrew-investigator`(定位代码)、`cavecrew-builder` (改 1-2 个文件)或 `cavecrew-reviewer`(审 diff),而不是自己 inline 做完 或用原版 `Explore`。Subagent 输出经过 caveman 压缩,塞回主 context 的 tool-result 小约 60%—— 长 session 下主 context 撑得更久。 Trigger: "delegate to subagent", "use cavecrew", "spawn investigator/builder/reviewer", "save context", "compressed agent output".

2 stars
0 votes
0 copies
0 views
Added 9/23/2026
ai-agents

Security Analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned 9/23/2026

$npx -y skills add Hayatelin/caveman-zh-CN --skill cavecrew --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Cavecrew?

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

Security grade badge for Cavecrew
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/hayatelin-cavecrew/badge)](https://www.skillsdirectory.com/skills/hayatelin-cavecrew)

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: cavecrew
description: >
  决定何时把工作委派给 caveman 风格 subagent 的判断指南。告诉主线程
  什么时候该派 `cavecrew-investigator`(定位代码)、`cavecrew-builder`
  (改 1-2 个文件)或 `cavecrew-reviewer`(审 diff),而不是自己 inline 做完
  或用原版 `Explore`。Subagent 输出经过 caveman 压缩,塞回主 context 的
  tool-result 小约 60%——
  长 session 下主 context 撑得更久。
  Trigger: "delegate to subagent", "use cavecrew", "spawn investigator/builder/reviewer",
  "save context", "compressed agent output".
---

Cavecrew = 三个会输出 caveman 文体的 subagent 缺省组态。工作内容跟 Anthropic 原生的一样(`Explore`、编辑型 agent、reviewer);差别在它们返回的 tool-result 是压缩过的,所以每次委派吃掉的主 context 更少。

## 什么时候用 cavecrew、什么时候用别的

| 任务 | 用 |
|---|---|
| “X 定义在哪/谁调用 Y/列出 Z 的所有用处” | `cavecrew-investigator` |
| 同上,但你还想要建议或架构评论 | `Explore`(原版) |
| 外科手术式编辑,≤2 个文件,范围明确 | `cavecrew-builder` |
| 新功能/3 个以上文件/跨层重构 | 主线程或 `feature-dev:code-architect` |
| 审 diff、branch 或单文件找 bug | `cavecrew-reviewer` |
| 深度 code review,要理由+替代方案 | `Code Reviewer`(原版) |
| 你已经知道答案的一行问题 | 主线程,不派 subagent |

判断法则:**如果你希望 subagent 的输出只有 1/3 的 token,选 cavecrew。如果你希望看到成篇散文,选原版。**

## 这东西为什么存在(真正的收益)

Subagent 的 tool result 会原封不动塞进主 context。一个原版 `Explore` 返回 2k token 的散文,就是每次吃掉主 context 预算 2k token。同一个发现交给 `cavecrew-investigator`,返回约 700 token。一个 session 里委派 20 次,这个差距就是“context 爆掉”跟“任务做完”的差别。

## 输出契约

主线程可以依赖每个 agent 的这些格式:

**`cavecrew-investigator`**
```
<Header>:
- path:line — `symbol` — short note
totals: <counts>.
```
或 `No match.` 一律文件路径在前、行号紧接、symbol 用反引号包住。可以安全地用 `path:\d+` grep。

**`cavecrew-builder`**
```
<path:line-range> — <change ≤10 words>.
verified: <re-read OK | mismatch @ path:line>.
```
或以下其中之一:`too-big.` / `needs-confirm.` / `ambiguous.` / `regressed.`(终止用的第一个 token)。

**`cavecrew-reviewer`**
```
path:line: <emoji> <severity>: <problem>. <fix>.
totals: N🔴 N🟡 N🔵 N❓
```
或 `No issues.` 发现项目依 文件 → 行号 递增排序。

## 串接模式

**定位 → 修 → 验**(最常用):
1. `cavecrew-investigator` 返回位置清单。
2. 主线程挑 1-2 个位置,把路径交给 `cavecrew-builder`。
3. `cavecrew-reviewer` 审计那份 diff。

**平行侦察**(调查范围很广时):
在同一则消息里同时派 2-3 个 `cavecrew-investigator`(不同角度:定义 vs 调用端 vs 测试)。在主线程汇整。

**单发编辑**(位置已经确定时):
跳过 investigator。直接把精确的 path:line 交给 `cavecrew-builder`。

## 不要做的事

- 还不知道是哪个文件时,别用 `cavecrew-builder`。先派 investigator,否则主线程光传脉络就把 token 吃光。
- 别为了 5 个文件的重构去串 `cavecrew-investigator → cavecrew-builder`。Builder 会回 `too-big.`,你白白浪费一轮。
- 别叫 `cavecrew-reviewer` 给“一般性建议”——它只反馈发现项目,不谈架构意见。那种需求用 `Code Reviewer`。
- 别期待散文。Cavecrew 的输出是结构化的,有时精简到近乎天书。如果要直接给人看,请自行改写。

## 自动切回清楚模式(继承)

Subagent 遇到安全性警告、不可逆操作的确认,以及任何碎片式表达可能被误读的输出时,会放下 caveman 改用正常英文。之后恢复 caveman。

Attribution

HayatelinHayatelin
View sourceSee grades on GitHubMore from Hayatelin →
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 →