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/ros-claw/rosclaw/agents-mdgit clone --depth 1 https://github.com/ros-claw/rosclawWrote 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/ros-claw/rosclaw/agents-md)<a href="https://agentmods.dev/instructions/ros-claw/rosclaw/agents-md"><img src="https://agentmods.dev/badge/instructions/ros-claw/rosclaw/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 | $0.01717 | $0.01717 |
| Opus 5 | $0.00859 | $0.00859 |
| Sonnet 5 | $0.00343 | $0.00343 |
| Haiku 4.5 | $0.00172 | $0.00172 |
Grade A, and why
rosclaw AGENTS.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 — 143 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ROSClaw Agent Instructions
ROSClaw is physical-AI runtime infrastructure. Treat robot, ROS, actuator, motor, and hardware commands as safety-sensitive. This file is safe for Codex, Claude Code, OpenClaw, and other agent frameworks that read project guidance.
Runtime boundary
- Project root:
. - Default robot: (none detected)
- MCP transport:
stdio - Pinned ROSClaw CLI:
.venv/bin/python -m rosclaw.entrypoint; use it instead of anotherrosclawonPATH. - One-line setup:
rosclaw agent install --project-root . --skip-secrets - Codex activation: trust this exact repository, then run
rosclaw agent doctor codex --project-root .. - Claude Code activation: approve the project
rosclawMCP server.
Tool policy
The ROSClaw MCP server exposes 22 read-only, body-context, simulation, validation, guarded-action, and emergency tools. It exposes no raw hardware primitive and gives the Agent no authority to approve its own REAL request.
| Tool | Safety level | Purpose |
|---|---|---|
get_robot_state |
S0 | Current body state and readiness |
list_skills |
S0 | Skills available to the runtime |
query_memory |
S0 | Retrieve similar past experiences |
validate_trajectory |
S2 | Plan validation, never real motion |
sandbox_run |
S1 | MuJoCo simulation preview only |
practice_query |
S0 | Query practice episodes |
emergency_stop |
S4 | Request daemon E-Stop; verify physical-stop evidence |
get_body_profile |
S0 | Static effective body profile |
get_body_state |
S0 | Body safety state and capability matrix |
list_body_capabilities |
S0 | Capabilities grouped by status |
query_body |
S0 | Answer questions about the current body |
validate_body_action |
S0 | Validate proposed body-level action |
get_calibration_status |
S0 | Calibration status for body components |
get_runtime_status |
S0 | rosclawd health and privilege-boundary status |
request_action |
S3 | Submit a guarded SHADOW or REAL action to rosclawd |
get_action_status |
S0 | Read daemon queue state and terminal receipt |
cancel_action |
S3 | Cancel queued work; active motion requires E-Stop |
get_product_status |
S0 | Canonical release and evidence boundary |
list_product_demos |
S0 | Official evidence-bearing simulation demos |
run_product_demo |
S1 | Run an official simulation and persist its receipt |
get_execution_receipt |
S0 | Read and integrity-check a receipt |
explain_execution |
S0 | Explain policy, execution, observation, and evidence |
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 · 143 lines · 1,717 tokens per session scan A 0a9f625f02b3
rosclaw AGENTS.md is an instructions file published in the GitHub repository ros-claw/rosclaw (191 stars, last pushed today), licensed MIT. It adds 1,717 tokens to every session, about $0.0086 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 2026-08-30.
Other instructions, from other repositories
clawrun CLAUDE.md
Instructions for clawrun-sh/clawrun, covering claude.md — clawrun, what this is, repo structure, package responsibilities and packages/cli (clawrun).
clawrun AGENTS.md
Instructions for clawrun-sh/clawrun, covering agents.md — agent interface & implementation guide, agent interface, registration pattern, existing implementations and zeroclaw (@clawrun/agent-zeroclaw).
CialloClaw AGENTS.md
AGENTS.md instructions for 1024XEngineer/CialloClaw, covering agents.md, 1. 项目共同语义, 2. 最小阅读规则, 3. 冲突处理优先级 and 4. 共享工程铁律.
hoop CLAUDE.md
Claude Code instructions for hoophq/hoop, covering claude.md, project overview, toolchain & prerequisites, architecture and module breakdown.
gini-agent AGENTS.md
AGENTS.md instructions for Open-Curiosity/gini-agent, covering gini agent instructions, shape, adrs, boundaries and branches.
autonomous-os AGENTS.md
AGENTS.md instructions for autonomous-ai/autonomous-os, covering agents.md, multi-ide rules, working style, parallel work / subagents and device access rules.