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/karanshukla/openresto/claude-mdgit clone --depth 1 https://github.com/karanshukla/openrestoWrote 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/karanshukla/openresto/claude-md)<a href="https://agentmods.dev/instructions/karanshukla/openresto/claude-md"><img src="https://agentmods.dev/badge/instructions/karanshukla/openresto/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 | $0.21238 | $0.21238 |
| Opus 5 | $0.10619 | $0.10619 |
| Sonnet 5 | $0.04248 | $0.04248 |
| Haiku 4.5 | $0.02124 | $0.02124 |
Grade A, and why
openresto 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 today.
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 — 385 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
Commands
Run everything (recommended for local dev)
npm run dev # starts backend (dotnet watch) + frontend (expo) concurrently
npm run dev:native # the same stack, addressed so a phone running Expo Go can reach it
npm test # the root dev scripts' own tests (node --test, no jest at this level)
Both pass --kill-others --kill-timeout to concurrently, so one half dying takes the other
down instead of leaving the backend squatting on :8080 where the next dotnet watch cannot
bind. concurrently kills through tree-kill, which is what reaches the
dotnet watch → dotnet run → OpenRestoApi grandchild that actually holds the port.
dev:native exists because openresto-frontend/.env points the app at http://localhost:8080,
and in Expo Go localhost is the phone, so the app resolves it to itself and reaches no API
at all. scripts/dev-native.mjs resolves this machine's LAN address and passes it as
EXPO_PUBLIC_API_URL, which Metro bakes into the bundle at build time; @expo/env skips any key
already in the system environment, so that wins over the .env file rather than racing it.
Docker and podman bridges are filtered out deliberately — bring the compose stack up and their
172.x/10.88.x addresses appear alongside the real one, and picking those fails exactly the way
localhost does. Override with OPENRESTO_DEV_HOST / OPENRESTO_API_PORT when the guess is
wrong; the script exits rather than falling back to an address the phone cannot route to.
Which address that is comes from openresto-frontend/scripts/lib/expo-go.mjs, shared with
npm run native:go rather than reimplemented — the two answer the same question, and answering
it twice is how the answers drift. It asks the routing table (a connected UDP socket fixes the
local end of a route without sending anything) and only falls back to filtering interface names.
OPENRESTO_DEV_HOST is that lib's HOST_ENV, so both commands take the same override; the
lib imports node builtins only, so the root npm test needs no frontend install.
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.
- today Changed · +1 lines · +274 tokens per session 23bec57d43fe
- yesterday Changed · +66 lines · +7,522 tokens per session 64ab8c056a1c
- 4d ago First seen · 318 lines · 13,442 tokens per session scan A d38af010ca3f
openresto CLAUDE.md is an instructions file published in the GitHub repository karanshukla/openresto (88 stars, last pushed today), licensed MIT. It adds 21,238 tokens to every session, about $0.1062 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 2026-09-01.
Other instructions, from other repositories
netrock CLAUDE.md
Claude Code instructions for fpindej/netrock, covering claude.md, hard rules, backend, frontend and cross-cutting.
swarmpit CLAUDE.md
Claude Code instructions for swarmpit/swarmpit, covering swarmpit, build / test / dev, source layout, domain model + non-obvious behaviour and frontend conventions.
Dotnet.Samples.AspNetCore.WebApi CLAUDE.md
Claude Code instructions for nanotaboada/Dotnet.Samples.AspNetCore.WebApi, covering claude.md, structure, coding guidelines, test naming conventions and fluentvalidation rule sets.
FamilyNido CLAUDE.md
Instructions for pablitofernandez/FamilyNido, covering claude.md, c# development section start, c# instructions, general instructions and naming conventions.
phleet CLAUDE.md
Claude Code instructions for anurmatov/phleet, covering fleet — autonomous ai agent system, repository structure, build commands, build the entire solution and build a specific project.
Coffer CLAUDE.md
Claude Code instructions for bartoszclapinski/Coffer, covering coffer, quick orientation, bieżąca praca — najpierw to, tech stack and project layout.