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 skills/autonomous-ai/autonomous-os/guardnpx skills add autonomous-ai/autonomous-os --skill guardgit clone --depth 1 https://github.com/autonomous-ai/autonomous-osWrote 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/skills/autonomous-ai/autonomous-os/guard)<a href="https://agentmods.dev/skills/autonomous-ai/autonomous-os/guard"><img src="https://agentmods.dev/badge/skills/autonomous-ai/autonomous-os/guard.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 5.1 | $0.00101 | $0.01815 |
| Opus 5 | $0.00051 | $0.00907 |
| Sonnet 5 | $0.00020 | $0.00363 |
| Haiku 4.5 | $0.00010 | $0.00181 |
Grade A, and why
guard scanned grade A 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 6d 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 URLlowData 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://127.0.0.1:5000/api/guard/enable -H 'Content-Type: application/json' -d '{"instruction":"custom instruction here"}' Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -s -X POST http://127.0.0.1:5000/api/guard/enable -H 'Content-Type: application/json' -d '{"instruction":"custom instruction here"}' How it starts
The opening of the file, as written. The whole thing — 107 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Guard Mode
Quick Start
Guard mode turns the device into an alert watchdog. When enabled, the device monitors for strangers and reacts dramatically — jolting, flashing, and verbally describing intruders. The system auto-broadcasts alerts to Telegram so friends know what's happening. All friends (enrolled faces) can toggle guard mode.
Workflow
- Friend requests guard mode (explicit or implied departure).
- Reply with
[HW:/emotion:{"emotion":"acknowledge","intensity":0.7}]— the device nods and flashes green. - Enable guard mode via the API (include
instructionif user gave extra requests). - Confirm verbally: "Guard mode on. I'll keep watch."
- While active: react dramatically to stranger/motion events (see trigger table below).
- When a friend returns, greet and ask if they want to disable. Only disable on explicit confirm.
Enable Guard Mode
curl -s -X POST http://127.0.0.1:5000/api/guard/enable -H 'Content-Type: application/json' -d '{"instruction":"custom instruction here"}'
The instruction field is optional. Use it when the user adds extra instructions on what to do during guard mode (e.g. "play scary sound when stranger detected", "flash red lights", "play alarm music"). Extract the relevant part from the user's message and pass it as the instruction. If the user just says "guard mode" with no extra requests, omit the field or send an empty body.
Response: {"status": 1, "data": {"guard_mode": true, "instruction": "..."}}
Disable Guard Mode
curl -s -X POST http://127.0.0.1:5000/api/guard/disable
Response: {"status": 1, "data": {"guard_mode": false}}
Check Guard Status
curl -s http://127.0.0.1:5000/api/guard
Response: {"status": 1, "data": {"guard_mode": true}}
Trigger Phrases
| User says | Action |
|---|---|
| "guard mode" / "watch mode" / "security mode" | Enable guard mode |
| "I'm going out" / "I'm leaving" / "bye, watch the house" | Enable guard mode |
| "stop guarding" / "I'm back" / "guard off" | Disable guard mode |
| "are you guarding?" / "guard status" | Check and report status |
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 6d ago First seen · 107 lines · 101 tokens per session scan A 1dd5333070c4
guard is a skill published in the GitHub repository autonomous-ai/autonomous-os (277 stars, last pushed yesterday), licensed Apache-2.0. It adds 101 tokens to every session and 1,815 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A 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 skills, from other repositories
rosclaw-simforge
Safely install, validate, diagnose, and optimize ROSClaw simulation workflows across MuJoCo/MJWarp, ROS 2 rosbridge, turtlesim, Gazebo, Isaac Sim, Isaac Lab multi-GPU training, MCP, and the signed ROSClaw Hub. Use for evidence-backed physical-AI smoke tests, 4-GPU validation, simulator integration, Hub upload/download…
rosclaw-embodied
Skill "rosclaw-embodied" from ros-claw/rosclaw, covering rosclaw 具身任务纪律, 权威资产, 证据与验收 and 安全分层.
provider-adapter-development
Use for WorldForge provider work: adding adapters, changing capability declarations, promoting scaffolds, debugging provider failures, updating provider catalog docs, or touching LeWorldModel, GR00T, LeRobot, Cosmos-Policy, JEPA, Genie, or JEPA-WMS. Ensures capabilities remain truthful and optional runtimes stay…
optional-runtime-smokes
Use for LeWorldModel, GR00T, LeRobot, PushT robotics showcase, real-checkpoint smoke scripts, checkpoint building, and host-owned optional runtime dependencies. Keeps real-runtime validation explicit without adding heavy ML/robotics packages or artifacts to the base project.
tui-development
Use for the robotics showcase Textual UI: report panes, launch helpers, screenshots, visual tests, and changes under src/worldforge/harness/tui.py or robotics view/rendering modules. Preserves the optional Textual boundary while keeping robotics flow logic testable without the TUI.
rosclaw
Use when operating, validating, or changing ROSClaw physical-AI runtime workflows, especially CLI smoke tests, Practice evidence loops, body/runtime checks, MCP integration, MuJoCo sandbox verification, and safe ROS or hardware boundaries.