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 skills add nebius/nebius-physical-ai --skill lichtblickgit clone --depth 1 https://github.com/nebius/nebius-physical-aiWrote 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/nebius/nebius-physical-ai/lichtblick)<a href="https://agentmods.dev/skills/nebius/nebius-physical-ai/lichtblick"><img src="https://agentmods.dev/badge/skills/nebius/nebius-physical-ai/lichtblick/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/nebius/nebius-physical-ai/lichtblick"><img src="https://agentmods.dev/badge/skills/nebius/nebius-physical-ai/lichtblick.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00041 | $0.01987 |
| Opus 5 | $0.00020 | $0.00993 |
| Sonnet 5 | $0.00008 | $0.00397 |
| Haiku 4.5 | $0.00004 | $0.00199 |
Grade A, and why
lichtblick 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 3d 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 — 124 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Lichtblick
Lichtblick is an open-source (MPL-2.0), Foxglove-compatible MCAP / ROS-bag /
robotics log viewer. It is the browser-based analog of rerun-viewer: a static
web viewer that opens artifacts, with no GPU. The image ships
Lichtblick (lichtblick-suite/lichtblick), the actively maintained community
fork of the archived, relicensed Foxglove Studio. It is a distinct product from
the now-proprietary Foxglove; no Foxglove account or proprietary component is
required.
- Tool name:
lichtblick— Image:npa-lichtblick— Port:8080— Tier:service. - Pinned OSS version: Lichtblick
1.26.0([tool.npa.supported-tools]). - CPU-only; not part of the
cuda12/cuda13-b300tag families.
Related:
skills/tools/foxglove/SKILL.mdcovers the official Foxglove app embedded with the@foxglove/embedSDK, the MCAP converter (npa workbench foxglove convert-run), and the agent's Foxglove pane — which falls back to this viewer when no Foxglove account is configured.
Interfaces
CLI:
# View an existing MCAP (plan only by default; --execute stages + launches):
npa workbench lichtblick serve --input-path s3://bucket/run/recording.mcap --execute
# Pack a robot camera-frame sequence (e.g. sim2real rollout/augment frames) into
# a real MCAP of foxglove.CompressedImage messages, then serve it:
npa workbench lichtblick serve \
--input-path s3://bucket/sim2real-b/<run-id>/augment/frames/ \
--from-frames --fps 10 --topic /sim2real/augment/camera --execute
npa workbench lichtblick launch # alias for serve
npa workbench lichtblick status
npa workbench lichtblick list
# Decode an MCAP into a native Rerun .rrd so Rerun renders it too (Rerun's built-in
# MCAP loader keeps foxglove/JSON messages as raw blobs; this maps the well-known
# schemas to archetypes). Open the result with `rerun x.rrd` or `sim2real rerun serve`.
npa workbench lichtblick to-rerun \
--input-path s3://bucket/sim2real-b/<run-id>/reports/sim2real.mcap \
--output-path <run-id>.rrd --execute
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.
- 3d ago Changed · +3 lines 09f826b84ce6
- 5d ago Changed 5c6281be8be0
- 11d ago First seen · 121 lines · 41 tokens per session scan A 7204c146298f
lichtblick is a skill published in the GitHub repository nebius/nebius-physical-ai (27 stars, last pushed today), licensed Apache-2.0. It adds 41 tokens to every session and 1,987 once invoked, about $0.0002 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 skills, from other repositories
screen-recording-on-chromebook
Records screen video on ChromeOS devices, Chromebooks, and CrOS DUTs over SSH using ffmpeg KMS grab, and transcodes the video to MP4 on the host workstation. Use when recording the screen of a Chromebook, ChromeOS device, cros DUT, or CrOS test hardware, capturing visual bug reproductions or feature demonstrations on…
physical-chromebook-debugging
Build and deploy Chrome to a physical ChromeOS device, monitor logs, and take screenshots over SSH. Use when working with remote Chromebook testing hardware, deploying new builds of Chrome or ChromeOS, reading specific log files via SSH, or taking screenshots of a connected test endpoint.
go2rtc-cameras
Multi-camera RTSP to WebRTC streaming via go2rtc.
cad-viewer-skill
Start or reuse CAD Viewer and return review links for CAD, implicit, robot-description, and G-code files (.step, .stl, .glb, .3mf, .gcode, .dxf, .urdf, .srdf, .sdf).
qualcomm-ai-assistant
A browser-based question tool for Qualcomm’s official product documentation and AI Assistant. It covers Qualcomm IoT chips and development boards such as IQ-9075, QCS9075, QCS6490, and QCS8250.
digikey
Proven scraping playbook for digikey.com /en/products/category/ listings. Cloudflare challenge (cf-mitigated: challenge) blocks bare curl with HTTP 403, but a puppeteer connection to the user's real Chrome via the real Chrome debug port walks straight through with zero stealth, no captcha, no proxy, no auth. Listings…