chamber CLAUDE.md

chamber CLAUDE.md is an instructions file for coding agents from abm9111/chamber. It costs 1,351 tokens per session, scanned C, original, MIT.

Project instructions for working on Chamber, a codebase whose important checks are called probes. Probes test real failure paths, while ordinary tests mainly check intended behavior with substitutes.

In plain words
What is it for?
Use them when changing Chamber to run probes, tests, type checks, and linting, and to preserve the project’s verification and safety rules.
Why use it?
They prevent a passing test suite from hiding failures in sandboxing, audit logs, checkpoints, debt handling, or other safeguards.

Instructions file

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 instructions/abm9111/chamber/claude-md
Clone the repo
git clone --depth 1 https://github.com/abm9111/chamber

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for chamber CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/abm9111/chamber/claude-md.svg)](https://agentmods.dev/instructions/abm9111/chamber/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/abm9111/chamber/claude-md"><img src="https://agentmods.dev/badge/instructions/abm9111/chamber/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 1,351 This file is loaded in full into every session.
When invoked 1,351 The same file — it is already loaded in full.
Security scan C 1 finding. 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 $0.01351 $0.01351
Opus 5 $0.00675 $0.00675
Sonnet 5 $0.00270 $0.00270
Haiku 4.5 $0.00135 $0.00135

Measured 4d ago against content hash 69f4100b9426, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

chamber CLAUDE.md scanned grade C with 1 finding 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 4d 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.

Reaches for credential fileshighPrivilege escalation

SSH keys, cloud credentials, git-credentials, .npmrc, /etc/shadow: reading these is how a config file becomes a credential leak.

`homedir()` does not (bwrap creates it), and `/etc/shadow` does not (EACCES either
CLAUDE.md · 104 lines

How it starts

The opening of the file, as written. The whole thing — 104 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Working on Chamber

Chamber's claim is that its gates are real. Everything below was learned by a gate turning out not to be — usually while a green test suite said otherwise.

The gates that matter are the probes, not the tests

npm test passed at 312, 315, 319, 324, 329, 340 and 350 tests while, in the same tree, the sandbox denied all execution on Linux, the debt gate parked every commit, an audit row could never be written, and a vendor tool description chose the code that ran. The unit suite uses stubs and encodes intent; it cannot fail for the reasons that matter.

npm run probes     # the gate. exits non-zero if any escape reproduces.
npm test           # necessary, never sufficient
npm run typecheck
npm run lint

CI gates on all four. A probe that cannot fail is not a gate — if you add one, make sure it can go red, and prove it once by breaking the thing it watches.

Verification rules that have each been paid for

Ask what state exists at the moment a guard fires. This one question found defects in the sandbox, the debt gate, the checkpoint, the anchor log and the turn path. chamber turn committed an observation and then discovered it could not reach a model. exportCheckpoint wrote the receipt and then let the anchor append refuse. A precondition checked after the write is not a precondition.

Isolation cannot be tested on macOS. There is no bwrap here, so CHAMBER_SANDBOX_REQUIRED=1 refuses for the right reason and for every wrong reason identically. A test asserting "it refuses" passes either way. Verify on Linux:

docker run --rm --privileged -e CHAMBER_SANDBOX_REQUIRED=1 -v "$PWD:/app:ro" node:24-slim \
  bash -c 'apt-get update -qq && apt-get install -y -qq bubblewrap && cd /app && node --experimental-strip-types -e "…"'

Plant a secret in $HOME and assert the payload cannot read it. --privileged is needed for bwrap's namespaces. Note that bwrap applies mount ops left to right: a --tmpfs over a parent placed after a --bind beneath it silently hides the bind.

Read the full file on GitHub · 104 lines

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. 4d ago First seen · 104 lines · 1,351 tokens per session scan C 69f4100b9426

Subscribe to this mod's changes

chamber CLAUDE.md is an instructions file published in the GitHub repository abm9111/chamber (3 stars, last pushed 14d ago), licensed MIT. It adds 1,351 tokens to every session, about $0.0068 per session on Opus 5. A static security scan graded it C with 1 finding (reaches for credential files). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other instructions, from other repositories

obsidian-llm-wiki AGENTS.md

AGENTS.md instructions for green-dalii/obsidian-llm-wiki, covering llm wiki plugin project development standards, 🛡️ six-gate quality closure, gate 1: five-gate automated, gate 2: no side effects and gate 3: no breaking changes.

green-dalii/obsidian-llm-wiki · 7,700 tokens

obsidian-llm-wiki CLAUDE.md

Claude Code instructions for green-dalii/obsidian-llm-wiki, a project described as: Karpathy's LLM Wiki implementation plugin for Obsidian - turns notes and PDFs into a linked, LLM-powered knowledge base with entity pages, concept pages, graph-powered Q&A, and local-first privacy.

green-dalii/obsidian-llm-wiki · 234 tokens

commonplace CLAUDE.md

Instructions for briannaworkman/commonplace: This repository holds two vaults. Nothing should be ingested or edited here at the root.

briannaworkman/commonplace · 192 tokens

CarpeOS AGENTS.md

AGENTS.md instructions for innocarpe/CarpeOS, covering carpeos agent guide, primary boundary, documentation rules, agent install (idempotent) and architecture rules.

innocarpe/CarpeOS · 1,988 tokens

Molio CLAUDE.md

Claude Code instructions for zhuzhaoyun/Molio, covering molio — 让你的积累,在 ai 时代持续产生价值, project structure (pnpm monorepo), build & dev commands, desktop 构建与运行 and docker / nas 一键部署.

zhuzhaoyun/Molio · 4,452 tokens

obsidian-mind CLAUDE.md

Claude Code instructions for breferrari/obsidian-mind, covering obsidian mind, skills & capabilities, custom slash commands, vault structure and obsidian cli.

breferrari/obsidian-mind · 9,405 tokens