caveman-help
Use when the user asks how caveman mode works, what intensities exist, what the slash commands are, or wants a quick-reference card before turning the mode on.
- License 许可
- MIT
- 最近更新
- 2026-11-09
- Stars
- ★ 6
- Forks
- 1
01 前言 Front Matter
- name
- caveman-help
- license
- MIT
- metadata
- { "author": "ai-playbook (ported from JuliusBrussee/caveman, MIT)", "version": "1.0" }
- description
- Use when the user asks how caveman mode works, what intensities exist, what the slash commands are, or wants a quick-reference card before turning the mode on.
03 SKILL.md 正文
caveman-help — quick-reference card
A one-shot reference card. Print and stop. Do not enter caveman mode just because this skill fires.
When to fire
User intent triggers (any of):
- "/caveman-help"
- "what is caveman", "how does caveman work", "explain caveman mode"
- "what caveman intensities are there", "caveman cheatsheet"
Output
Print exactly the block below, verbatim, then stop. Do not add commentary, do not enter caveman mode, do not adjust based on the user's tone.
caveman mode — quick reference
WHAT
Compress agent output ~65-75%. Brain unchanged. Mouth smaller.
Only affects output tokens, not thinking tokens.
INTENSITIES
lite — drop filler/hedging, keep articles + sentences (~25% saving)
full — drop articles, fragments OK (~65% saving, default)
ultra — abbreviate prose, arrows for causality (~80% saving)
SLASH COMMANDS
/caveman [mode] switch mode for this session
/caveman-compress F shrink markdown file F in place (backup at F.original.md)
/caveman-commit terse Conventional Commit subject
/caveman-review one-line PR review w/ severity emoji
/caveman-help this card
PROJECT TOGGLE (persistent across sessions)
python -m scripts.caveman status [--json]
python -m scripts.caveman on --mode full --components response_style,mcp_shrink
python -m scripts.caveman off [--keep-backups]
python -m scripts.caveman rollback
STOP TRIGGERS
"stop caveman", "normal mode", "be verbose", "/caveman off"
AUTO-CLARITY (mode pauses automatically for)
security warnings · irreversible actions · multi-step sequences · confused user
NOT COMPRESSED
code blocks · file paths · URLs · tool inputs · commit/PR bodies (unless opted in)
DOCS
docs/concepts/caveman-mode.md — what & why
docs/runbooks/caveman-toggle.md — how to use
docs/operations/caveman-architecture.md — UI contract
Trust boundary
- Do not modify the block. It is a single source of truth — change the SKILL.md if the reference needs updating.
- Do not flip into caveman style after printing. This skill is a help screen, not a mode switch.
- If the user follows up with "turn it on", do not act on it from this skill — direct them to
python -m scripts.caveman onor/caveman.
04 同仓库更多 Skills
sweep
Use when the user wants to find code that has quietly stopped being used — types /sweep, asks for a "dead code sweep", "repo entropy", "what can we delete", "orphan files", "unused code", or asks to review a sweep ledger. Adjudicates candidates a deterministic scanner already produced; it never deletes anything and never re-derives findings from raw source. Pairs with ponytail, which gates what gets written rather than what rots.
ponytail
Use when the user wants the laziest solution that actually works — types /ponytail, says "be lazy", "lazy mode", "simplest solution", "minimal solution", "yagni", "do less", or "shortest path", or complains about over-engineering, bloat, boilerplate, or unnecessary dependencies. Supports intensity levels lite, full, ultra (default). Pairs with caveman, which compresses prose, not code.
ponytail-review
Use when the user wants a code review focused exclusively on over-engineering — says "review for over-engineering", "what can we delete", "is this over-engineered", "simplify review", or invokes /ponytail-review. Complements correctness/security review; this one only hunts complexity in the current diff. Gated by the ponytail review_ponytail component.
ponytail-debt
Use when the user wants to harvest every `ponytail:` comment in the codebase into a tracked debt ledger — says "ponytail debt", "/ponytail-debt", "what did ponytail defer", "list the shortcuts", "ponytail ledger", or "what did we mark to do later". One-shot report, changes nothing. Gated by the ponytail debt_ponytail component.
ponytail-audit
Use when the user wants a whole-repo audit for over-engineering — says "audit this codebase", "audit for over-engineering", "what can I delete from this repo", "find bloat", or invokes /ponytail-audit. Like ponytail-review but scans the entire tree, not a diff. One-shot report, applies no fixes. Gated by the ponytail audit_ponytail component.
openspec-propose
Propose a new change with all artifacts generated in one step. Use when the user wants to quickly describe what they want to build and get a complete proposal with design, specs, and tasks ready for implementation.