CA

caveman-sonata

A terse mode for code implementation and progress updates that preserves harness engineering.

Install

mkdir -p .claude/skills/caveman-sonata && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/12573" && unzip -o skill.zip -d .claude/skills/caveman-sonata && rm skill.zip

Installs to .claude/skills/caveman-sonata

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Use when: /caveman-sonata, caveman mode, less tokens, terse implementation, adding modules, compact progress updates, or token-cheap harness-aware coding. Always preserve harness engineering.
191 chars✓ has a “when” trigger
Intermediate

Key capabilities

  • Implement tasks or features with terse code
  • Add new modules to a project
  • Provide compact progress updates
  • Preserve harness engineering flow
  • Create or update execution plans for broad work

How it works

The skill implements tasks tersely while adhering to a harness-aware flow, providing compact updates and exact technical details.

Inputs & outputs

You give it
Task or feature to implement tersely
You get back
Code, commands, paths, errors, and a compact summary with checks run

When to use caveman-sonata

  • Implement features with low token usage
  • Add code modules
  • Provide concise status updates

About this skill

Caveman Sonata

Terse like caveman. Harness-aware. Technical meaning exact.

Rules

  • Default mode: full.
  • Read only needed docs from docs/index.md.
  • Never skip harness flow: acceptance, plan when needed, implementation, validation, doc updates.
  • For broad work, create or update execution plan first.
  • Progress updates: short fragments.
  • Code, commands, paths, errors: exact.
  • Final answer: compact summary plus checks run.
  • Safety, destructive ops, confusing multi-step instructions: normal prose.

Pattern

Goal:
Change:
Check:
Next:

When not to use it

  • When safety, destructive operations, or confusing multi-step instructions are involved
  • When a verbose implementation or detailed progress updates are required

Limitations

  • Requires adherence to harness flow (acceptance, plan, implementation, validation, doc updates)
  • Progress updates must be short fragments
  • Final answer must be a compact summary plus checks run

How it compares

This skill enforces a terse, harness-aware implementation style with compact updates, unlike a verbose or less structured coding approach.

Compared to similar skills

caveman-sonata side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
caveman-sonata (this skill)03moNo flagsIntermediate
command-development169moReviewIntermediate
skill-forge119moReviewIntermediate
codex-skill125moReviewAdvanced

Try saying

Example prompts that trigger this skill in your AI assistant.

You might also like

command-development

anthropics

This skill should be used when the user asks to "create a slash command", "add a command", "write a custom command", "define command arguments", "use command frontmatter", "organize commands", "create command with file references", "interactive command", "use AskUserQuestion in command", or needs guidance on slash command structure, YAML frontmatter fields, dynamic arguments, bash execution in commands, user interaction patterns, or command development best practices for Claude Code.

16133

skill-forge

WilliamSaysX

Automated skill creation workshop with intelligent source detection, smart path management, and end-to-end workflow automation. This skill should be used when users want to create a new skill or convert external resources (GitHub repositories, online documentation, or local directories) into a skill. Automatically fetches, organizes, and packages skills with proactive cleanup management.

11115

codex-skill

feiskyer

Use when user asks to leverage codex, gpt-5, or gpt-5.1 to implement something (usually implement a plan or feature designed by Claude). Provides non-interactive automation mode for hands-off task execution without approval prompts.

12110

agent-factory

alirezarezvani

Claude Code agent generation system that creates custom agents and sub-agents with enhanced YAML frontmatter, tool access patterns, and MCP integration support following proven production patterns

8109

subagent-driven-development

davila7

Use when executing implementation plans with independent tasks in the current session

1493

peekaboo

openclaw

Capture and automate macOS UI with the Peekaboo CLI.

1486

Search skills

Search the agent skills registry