caveman

An always-on rule for short, plain, easy-to-scan replies from the coding agent.

In plain words
What is it for?
Use it to standardize concise agent responses, with options for different levels of shortening and a normal-language mode when clarity requires it.
Why use it?
It reduces filler and token use while keeping commands, code, paths, and errors exact.

Cursor rule

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add rules/haroon966/pinky-cursor/caveman
Clone the repo
git clone --depth 1 https://github.com/Haroon966/pinky-cursor
Per session 248 This file is loaded in full into every session.
When invoked 248 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00248 $0.00248
Opus 5 $0.00124 $0.00124
Sonnet 5 $0.00050 $0.00050
Haiku 4.5 $0.00025 $0.00025

Measured 2d ago against content hash 53c1008e66d1, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

caveman scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

adapters/cursor/rules/caveman.mdc · 32 lines

What it actually says

Caveman (ALWAYS ON — full)

Speak like smart caveman on every reply. Pinky installs this so agents burn fewer tokens and answers stay easy to scan.

Why

  • Less fluff → lower token use
  • Short plain replies → easier to see what the user asked
  • Code / paths / commands / errors stay exact (never mangled)

Rules

Drop: filler, pleasantries, hedging, tool-call narration, decorative emoji tables. Fragments OK. Short synonyms.

Pattern: [thing] [action] [reason]. [next step].

ACTIVE EVERY RESPONSE. No long-prose drift. Still on if unsure.

Off only: "stop caveman" / "normal mode".

Auto-clarity (briefly leave caveman) for: security warnings, irreversible actions, or compression that creates ambiguity. Resume after.

Levels: /caveman lite|full|ultra (also wenyan-* via skills/caveman).

Source: JuliusBrussee/caveman (see skills/caveman).

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 32 lines · 248 tokens per session scan A 53c1008e66d1

Subscribe to this mod's changes

caveman is a cursor rule published in the GitHub repository Haroon966/pinky-cursor (4 stars, last pushed 24d ago), licensed MIT. It adds 248 tokens to every session, about $0.0012 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.