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.
npx agentmods add instructions/juliusbrussee/caveman/claude-mdgit clone --depth 1 https://github.com/JuliusBrussee/cavemanWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.09291 | $0.09291 |
| Opus 5 | $0.04646 | $0.04646 |
| Sonnet 5 | $0.01858 | $0.01858 |
| Haiku 4.5 | $0.00929 | $0.00929 |
Grade A, and why
caveman CLAUDE.md scanned grade A with 2 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 yesterday.
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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
The `install.sh` / `install.ps1` shims at the repo root delegate to `bin/install.js` via `node` (local clone) or `npx -y github:JuliusBrussee/caveman` (curl|bash). No legacy fallback path remains — earlier `install.sh.le Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
**The stdin contour is load-bearing.** Activation runs on the first COMPLETE JSON object, not at EOF, with a 2000ms `PAYLOAD_WATCHDOG_MS` backstop and `process.stdin.unref()` in `finish()`. `pause()` alone is not enough The source is not reproduced here
No licence file
A repository with no LICENSE is all rights reserved by default, so the body is not copied here. The metadata, the measurements and the link are.
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.
- yesterday First seen · 416 lines · 9,291 tokens per session scan A ce0a6ca0e980
caveman CLAUDE.md is an instructions file published in the GitHub repository JuliusBrussee/caveman (101,863 stars, last pushed 2d ago), with no licence file. It adds 9,291 tokens to every session, about $0.0465 per session on Opus 5. A static security scan graded it A with 2 findings (makes network calls, runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
cavemem CLAUDE.md
Instructions for JuliusBrussee/cavemem, covering cavemem — agent playbook, project identity, non-negotiable rules, architectural rules and layout.
genshijin copilot-instructions.md
Instructions for InterfaceX-co-jp/genshijin: 原始人のように簡潔に返答せよ。技術的中身はすべて残す。無駄だけ消す。日本語前提。.
genshijin AGENTS.md
Instructions for InterfaceX-co-jp/genshijin: @./skills/genshijin/SKILL.md @./skills/genshijin-commit/SKILL.md @./skills/genshijin-review/SKILL.md @./skills/genshijin-help/SKILL.md @./skills/genshijin-compress/SKILL.md @./skills/genshijin-stats/SKILL.md @./skills/genshijin-crew/SKILL.md.
my-adhd-caveman AGENTS.md
Instructions for SteinsFu/my-adhd-caveman, covering agent guide, start here, repository map, source-of-truth rules and install for the user.
my-adhd-caveman GEMINI.md
Instructions for SteinsFu/my-adhd-caveman: ADHD shape + caveman diction. Follow the skill below in full: action first, numbered terse steps, one next action, drop fluff.
cavegemma AGENTS.md
AGENTS.md instructions for JuliusBrussee/cavegemma, covering agents.md, what this project is, what "caveman" means in code, pipeline at a glance and specific traps (each cost real hours / $).