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/darkroomengineering/cc-settings/agents-mdgit clone --depth 1 https://github.com/darkroomengineering/cc-settingsWhat 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 | $0.03506 | $0.03506 |
| Opus 5 | $0.01753 | $0.01753 |
| Sonnet 5 | $0.00701 | $0.00701 |
| Haiku 4.5 | $0.00351 | $0.00351 |
Grade A, and why
cc-settings AGENTS.md 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.
How it starts
The opening of the file, as written. The whole thing — 349 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Darkroom Engineering
Portable coding standards for Claude Code, Codex, Cursor, Copilot, Windsurf, and other AGENTS.md-compatible tools.
Philosophy
Make the codebase legible to agents and humans through written conventions, rules, and intent.
Getting Started
- Read this file.
- Read files, search code, and run builds directly.
- Delegate when the host's rules require it. Claude Code's thresholds live in
CLAUDE-FULL.md; multi-file exploration, security-sensitive code, and test writing MUST be delegated there. - Start simple and add complexity only when needed.
Response Calibration
Match length to the request. Lead with the answer. A lookup gets a sentence and file:line; a multi-file change gets a brief plan and landing summary. Skip preambles and duplicate recaps.
Register
Optimize for first-read comprehension, not brevity:
- Put the subject first; move modifiers into a clause instead of stacking them.
- Prefer active voice and direct verbs. Name the actor when it matters.
- Keep one subject or decision per sentence when combining claims creates ambiguity.
- Use an existing name for an existing thing; never coin or capitalize a replacement concept.
- Use identifiers when pointing at code, plain names when describing behavior.
- Define jargon inline on first use.
- State the effect before the mechanism. Never clip sentences or remove substance to be shorter.
This is the portable copy for AGENTS.md-aware tools and humans. Claude Code does not auto-load it: its subagents inherit the CLAUDE.md hierarchy, and its main sessions receive the same register from the Darkroom output style. Keep those delivery mechanisms distinct.
Guardrails
These rules are non-negotiable.
Laziness Ladder (Before Writing Code)
Stop at the first rung that holds:
- Does this need to exist? If not, skip it (YAGNI).
- Does this codebase already do it? Reuse or extend it.
- Does the standard library/runtime do it? Use it.
- Does the native platform do it? Use it.
- Does an installed dependency do it? Use it; add nothing.
- Can it be one line? Make it one line.
- Only then, write the minimum that works.
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.
- 2d ago First seen · 349 lines · 3,506 tokens per session scan A 76ccde045842
cc-settings AGENTS.md is an instructions file published in the GitHub repository darkroomengineering/cc-settings (42 stars, last pushed 4d ago), licensed MIT. It adds 3,506 tokens to every session, about $0.0175 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 2026-08-30.
Other instructions, from other repositories
research-os AGENTS.md
Instructions for lxinfei5/research-os, covering researchos — constitution, §0 what this is, §1 directory discipline, §2 half-life (innovation 1) — memory design and §3 behavior pillars (innovations 2–5).
agentic-workflow CLAUDE.md
Instructions for gtrabanco/agentic-workflow, covering claude.md, repository layout, working rules, authoring a skill and hand off, don't compose across a model/effort boundary.
skills CLAUDE.md
Instructions for capawesome-team/skills, covering claude.md, project overview, repository structure, skill.md format and writing agent-optimized skills.
plexe AGENTS.md
Instructions for plexe-ai/plexe, covering agents.md, architecture, 6-phase workflow (plexe/workflow.py), multi-agent system (plexe/agents/) and tree search (plexe/search/).
aeon CLAUDE.md
Instructions for aeonfun/aeon, covering aeon, how aeon works, strategy, voice and soul file hierarchy (read in this order).
best-of-Agent-Harnesses CLAUDE.md
Claude Code instructions for RyanAlberts/best-of-Agent-Harnesses, covering repo notes for ai coding agents, commit identity (required), branching and pushing (required), verifying contribution credit (required) and how the list is generated.