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/omnilink-tech/omnisim/agents-mdgit clone --depth 1 https://github.com/omnilink-tech/omnisimWrote 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/omnilink-tech/omnisim/agents-md)<a href="https://agentmods.dev/instructions/omnilink-tech/omnisim/agents-md"><img src="https://agentmods.dev/badge/instructions/omnilink-tech/omnisim/agents-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 | $0.49325 | $0.49325 |
| Opus 5 | $0.24663 | $0.24663 |
| Sonnet 5 | $0.09865 | $0.09865 |
| Haiku 4.5 | $0.04933 | $0.04933 |
Grade B, and why
omnisim AGENTS.md scanned grade B with 2 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.
Sends data to an external URLmediumData exfiltration
A POST to an outside endpoint may be telemetry or may be exfiltration; either way the mod talks to somewhere, and you should know where.
curl -s -X POST http://127.0.0.1:6789/world/screenshot -d '{}' -o shot.png Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- **⚠️ ROS 2 SHIPS — the "declared NON-GOAL" line this bullet used to carry was REVERSED on 2026-08-17, so do not answer from an older session's memory.** OmniSim implements the ROS 2 [`simulation_interfaces`](https://gi How it starts
The opening of the file, as written. The whole thing — 699 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Running OmniSim as an AI Coding Agent
This file is for AI coding agents (Claude Code, Codex, Cursor, custom agent harnesses) working inside a fresh clone of OmniSim. It tells you exactly what to do to get a demo running. The conventions below match the AGENTS.md open standard, and tools that respect that standard will load this file automatically.
If you are a human, this file also works as a quick "what does OmniSim do and how do I run a demo" cheat sheet — the README and the Developer Quickstart have the long form.
0. For agents: read this first (60 seconds)
Call it OmniSim. The simulator is OmniSim — its own product, with substantial additions over the upstream Webots engine it forked from (URDF importer, agent-facing HTTP harness, capture / cinema pipeline, OmniLink agent runtime, omniworld procedural world generator, RL pipeline, CUDA granular-physics, multi-instance parallel runs, OmniSim Wire Protocol). When you talk to the user about the running simulator, the binary, the env vars, the URL scheme — say OmniSim. Use "Webots" only when explicitly referring to upstream (the GitHub repo, the file-format syntax inherited from VRML, the original PROTO conventions).
A scenario = one .wbt world + one or more controllers + (optionally) a long-running harness or bridge service that you, the agent, drive over HTTP. You drive the simulator — you do not need to ask the user to launch, reload, or step it.
Run this first turn
python -m omnisim doctor
Reports the truth about this clone right now: OmniSim binary path, whether there is a physics runtime at all, engine↔libController ABI compatibility (the IPC-nonce gate, commit 6eea9d76 — a controller lib older than the engine silently hangs every controller at zero ticks while a headless run still prints PASS), port status (6789 harness, 6790 supervisor, 6791 capture), worlds present, recent commits. Don't guess at the state — check it. --json for machine-readable.
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 · 699 lines · 49,325 tokens per session scan B ea6172f6c16e
omnisim AGENTS.md is an instructions file published in the GitHub repository omnilink-tech/omnisim (82 stars, last pushed 4d ago), licensed Apache-2.0. It adds 49,325 tokens to every session, about $0.2466 per session on Opus 5. A static security scan graded it B with 2 findings (sends data to an external url, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
mjswan AGENTS.md
AGENTS.md instructions for ttktjmt/mjswan, covering agents.md, project, philosophy, layout and python workflow.
AReaL AGENTS.md
AGENTS.md instructions for areal-project/AReaL, covering agents.md -- areal agent operations guide, quick reference, environment, tests and cli docs.
AReaL CLAUDE.md
Claude Code instructions for areal-project/AReaL, covering claude.md - areal, what: project overview, why: purpose, how: core commands and check environment.
Gym AGENTS.md
AGENTS.md instructions for NVIDIA-NeMo/Gym, covering agents.md, quality bar, what this is, architecture and creating environments.
chrono AGENTS.md
AGENTS.md instructions for projectchrono/chrono, covering repository guidelines, developers vs. users, project structure & module organization, build, test, and development commands and build system conventions.
mobilegym AGENTS.md
Instructions for Purewhiter/mobilegym, covering agents.md, project overview, type-checking strategy, eslint and with data graph generation.