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/goplasmatic/orion/claude-mdgit clone --depth 1 https://github.com/GoPlasmatic/OrionWrote 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/goplasmatic/orion/claude-md)<a href="https://agentmods.dev/instructions/goplasmatic/orion/claude-md"><img src="https://agentmods.dev/badge/instructions/goplasmatic/orion/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 | $0.08115 | $0.08115 |
| Opus 5 | $0.04057 | $0.04057 |
| Sonnet 5 | $0.01623 | $0.01623 |
| Haiku 4.5 | $0.00812 | $0.00812 |
Grade A, and why
Orion 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.
| A route, method or query parameter | the book's `curl` examples + `docs/openapi.json` (`docs_routes_drift_test`, `openapi_test`) | How it starts
The opening of the file, as written. The whole thing — 251 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.
Project Overview
Orion is a declarative services runtime written in Rust. It exposes business logic management through channels (service endpoints) and workflows (task pipelines powered by dataflow-rs) via a REST API. Ships as a single binary with an embedded SQLite database.
This repo is a cargo workspace (monorepo) with four crates: crates/orion-server (the runtime — everything below describes it), crates/orion-cli (the CLI, which has its own CLAUDE.md), and two shared library crates. crates/orion-api is the wire contract: response DTOs, domain enums, the error envelope + codes registry, and the bulk-import report — the server serializes these types, clients deserialize them; the server re-exports them under their pre-1.0 paths (e.g. crate::errors::FieldError, storage::models::EntityStatus). Its deserialization is tolerant by design (every field defaults, unknown fields ignored) so version skew between server and CLI keeps working; its utoipa feature (enabled only by the server) adds the ToSchema derives for the OpenAPI document. crates/orion-client is the one HTTP transport over that contract — OrionClient (auth, envelope unwrap, typed ClientError) plus the paths module every endpoint path is built from; both orion-cli and the server's package_cli drive it, and it is the only crate that owns reqwest for API calls. Neither library crate has a release cycle of its own — no tag names them; crates-publish.yml publishes them automatically as riders (skip-if-present, dependency order) right before orion-server, because crates.io refuses a crate whose dependency it doesn't host. All four crates share one version, workspace.package.version in the root Cargo.toml, inherited with version.workspace = true and repeated only in the two [workspace.dependencies] requirements (a path dependency needs a version to be publishable). That is what makes the riders safe: a release always moves the number, so skip-if-present can never fire on a rider whose contents changed. Bump it with cargo release, never by hand in a member manifest. Note that orion-server is the only binary crate on crates.io: the orion-cli name there belongs to an unrelated 2021 crate, so the CLI ships via the dist installers, the Homebrew tap, GHCR and cargo install --git instead (see RELEASING.md). default-members points bare cargo commands at the server; use --workspace or -p orion-cli to reach the CLI. Only the UI lives in a separate repo (Orion-ui).
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 · 251 lines · 8,115 tokens per session scan A d13598e94cfe
Orion CLAUDE.md is an instructions file published in the GitHub repository GoPlasmatic/Orion (9 stars, last pushed 5d ago), licensed Apache-2.0. It adds 8,115 tokens to every session, about $0.0406 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
rocketride-server AGENTS.md
AGENTS.md instructions for rocketride-org/rocketride-server, covering agents.md and co-located documentation rule.
genblaze AGENTS.md
AGENTS.md instructions for backblaze-labs/genblaze, covering agents, repo purpose, architecture boundaries, invariants and guardrails and doc map.
genblaze CLAUDE.md
Claude Code instructions for backblaze-labs/genblaze, a project described as: Genblaze is an open source Python SDK for orchestrating generative AI media pipelines across video, audio, and image providers with built in provenance for every output.
traverse CLAUDE.md
Instructions for traverse-framework/traverse, covering traverse development guidelines, governance, active technologies, project structure and commands.
traverse AGENTS.md
Instructions for traverse-framework/traverse, covering traverse — agent coordination, agent coordination, 1. check for claude code claim, 2. check for claude code branch and 3. claim the ticket (only if pre-flight passes).
dagu CLAUDE.md
Claude Code instructions for dagucloud/dagu, covering dagu repository guide, project, repository map, runtime flow and architecture rules.