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/hariharapanigrahy/layerkit/claude-mdgit clone --depth 1 https://github.com/hariharapanigrahy/layerkitWrote 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/hariharapanigrahy/layerkit/claude-md)<a href="https://agentmods.dev/instructions/hariharapanigrahy/layerkit/claude-md"><img src="https://agentmods.dev/badge/instructions/hariharapanigrahy/layerkit/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 | $0.00385 | $0.00385 |
| Opus 5 | $0.00192 | $0.00192 |
| Sonnet 5 | $0.00077 | $0.00077 |
| Haiku 4.5 | $0.00038 | $0.00038 |
Grade A, and why
layerkit CLAUDE.md scanned grade A 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 5d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
Evidence-first (OpenAPI, docs, curl, code); residual human only; no LLM on track(). This is a copy
88% identical to layerkit AGENTS.md — 20 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
What it actually says
Layerkit
Evidence-first (OpenAPI, docs, curl, code); residual human only; no LLM on track().
Intentional entry only. Use layerkit: … (or /layerkit / @layerkit) for vendor integrate / contract heal, then layerkit help → agent start → agent next (skill packet). Do not freestyle pin-only full integrates. Do not apply Layerkit rails to unrelated tasks.
Master skill: layerkit-orchestrate-integration. Multi-agent coordination lives in the layerkit-multi-agent skill; there is no deterministic agent multi CLI.
Contract update: use layerkit agent start --mode heal --vendor …, research evidence, then edit client source/tests directly with layerkit-source-edit-client.
Checker-assist is read-only. Promote only after quality gates.
Before implementation starts, plans must name outcome checkpoints: what must pass, what artifact proves it passed, and the fallback if the design does not validate. For every strategic redirect or large rewrite/deletion, define the proof first: passing judge, package fixture, release checklist item, or concrete before/after behavior.
Keep test backing proportional to implementation size. Release hardening work must include executable coverage for client-package edit paths, mapping semantics, deletion-first behavior, and CI/eval gates. For public/shared repos, hardcoded API keys, passwords, tokens, and credentials are release blockers; move them to environment variables or a secrets manager.
Collaboration style: work with the user in a "Dances with Robots" rhythm: evidence-led, iterative, and explicit about tradeoffs as ideas bounce back and forth.
Shared AI working rules: AI_WORKING_RULES.md.
Docs: layerkit cheatsheet, skills/*/SKILL.md. CLI: layerkit / npx layerkit.
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.
- 5d ago First seen · 22 lines · 385 tokens per session scan A 33efd8b072d3
layerkit CLAUDE.md is an instructions file published in the GitHub repository hariharapanigrahy/layerkit (8 stars, last pushed 28d ago), licensed MIT. It adds 385 tokens to every session, about $0.0019 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). It is 88% identical to layerkit AGENTS.md, differing in 20 lines, and is treated as a copy.
Other instructions, from other repositories
the-loop CLAUDE.md
Claude Code instructions for MadaraUchiha-314/the-loop, covering claude.md — working in the-loop's own repository, why this file exists and what to do.
cdv copilot-instructions.md
Instructions for azank1/cdv, covering cdv — agent instructions, required on every message and memory & recovery (v0.8).
repo-ready AGENTS.md
Instructions for shidesheng0218/repo-ready, covering agents.md, project overview, common commands, agent rules and 中文说明.
repo-ready CLAUDE.md
Instructions for shidesheng0218/repo-ready, covering claude.md, stack, commands and rules.
agentic-harness AGENTS.md
Instructions for sevenschulte/agentic-harness, covering agents.md — how this system works, identity, stack, architecture (one paragraph) and repo layout.
breaking-coding-chaos AGENTS.md
AGENTS.md instructions for bo-cao/breaking-coding-chaos, a project described as: A human-in-the-loop control plane for reliable agentic coding—plan, challenge, implement, verify, and preserve progress.