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/hautc-it/cil/claude-mdgit clone --depth 1 https://github.com/hautc-it/cilWrote 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.
[](https://agentmods.dev/instructions/hautc-it/cil/claude-md)<a href="https://agentmods.dev/instructions/hautc-it/cil/claude-md"><img src="https://agentmods.dev/badge/instructions/hautc-it/cil/claude-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00930 | $0.00930 |
| Opus 5 | $0.00465 | $0.00465 |
| Sonnet 5 | $0.00186 | $0.00186 |
| Haiku 4.5 | $0.00093 | $0.00093 |
Grade A, and why
cil CLAUDE.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 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.
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 — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Response Style
Ask before proceeding when: requirements are ambiguous, scope expands beyond what was asked, a change is irreversible, or touching security/auth/data migration. Proceed autonomously when: the task is clearly scoped, tests will catch regressions, and the change is reversible.
Default: concise fragments. Pattern: [thing] [action] [reason].
Mid-task checkpoints (terse): Step 2/5: schema migrated. Running tests.
Explanations/design (prose): full sentences, grammar preserved.
Use prose for: security warnings, ambiguous sequences, error messages.
Output Compression
The PreToolUse hook auto-rewrites verbose commands (git diff/log/show, npm install/test, pytest, docker, cargo, tsc, pip install, find, ls -la). For commands the hook doesn't catch, manually append 2>&1 | cil compress when output is likely > 50 lines.
Modes: --mode=lite (filter only) | full (default — dedupe + group + truncate) | ultra (+ semantic dedup, best for test runners).
Code
Write the minimum code that solves the stated problem. No speculative features. Touch only what's asked. No opportunistic refactors. Validate at boundaries (user input, external APIs). Trust internal code. Fail fast with explicit errors. No silent failures, no swallowed exceptions. Every bug fix ships with a test that would have caught it.
Before implementing:
- Read relevant files. Map what exists.
- State assumptions explicitly.
- Define verifiable success criteria.
Security
Never log or expose: secrets, tokens, passwords, PII. Validate and sanitize all external input before use. Parameterize all queries. No string interpolation into SQL/shell. Flag any change touching auth, permissions, or data migration before proceeding.
Context Engineering — Write / Select / Compress / Isolate
CIL implements all four. Reach for them in this order:
| Operation | When | Primitive |
|---|---|---|
| Write | A decision, constraint, or learning is worth keeping for the next session | memory_store(category, content, tags[]) via MCP, or /learn <insight> |
| Select | Starting a new task — retrieve before re-deriving | memory_search(query) (auto-run by /develop Phase 1, including feedback-tagged corrections) |
| Compress | A command's output is likely > 50 lines | cil compress (auto via PreToolUse hook for known verbose commands) |
| Isolate | Two parallel tracks of work that must not contaminate each other's context | Claude Code's native worktree (EnterWorktree / ExitWorktree) — CIL does not duplicate this |
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.
- 4d ago First seen · 83 lines · 930 tokens per session scan A 6fd11b2925c7
cil CLAUDE.md is an instructions file published in the GitHub repository hautc-it/cil (1 stars, last pushed 1mo ago), licensed MIT. It adds 930 tokens to every session, about $0.0047 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.
Other instructions, from other repositories
embedded-systems-architect AGENTS.md
Instructions for Zachariah9420/embedded-systems-architect: This repository ships an agent skill for embedded-systems and edge-AI development work.
pbnz-skills AGENTS.md
AGENTS.md instructions for PBNZ/pbnz-skills, covering agents.md — pbnz-skills, start-here map — where things live and ground rules.
qa-orchestra CLAUDE.md
Claude Code instructions for Anasss/qa-orchestra, covering qa orchestra — claude code instructions, what this project is, quick start, 1. clone into your workspace and 2. fill in your project context.
agent-skills CLAUDE.md
Instructions for mindbox-cloud/agent-skills, covering agent-skills — developer guide for claude code, repository purpose, repository structure, how to add a new plugin and step 1 — create the directory structure.
claude-sdlc CLAUDE.md
Claude Code instructions for lantisprime/claude-sdlc, covering claude.md, what this repo is, design intent — read this before "improving" anything, eat your own dog food and hook strictness philosophy.
claude-account-switch AGENTS.md
AGENTS.md instructions for LightZirconite/claude-account-switch, covering claude + codex account switch - agent guide, product invariants, authentication ux, switching safety and provider parity and deliberate differences.