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 Mondlak1/camclave --skill skillgit clone --depth 1 https://github.com/Mondlak1/camclaveWrote 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/mondlak1/camclave/skill)<a href="https://agentmods.dev/skills/mondlak1/camclave/skill"><img src="https://agentmods.dev/badge/skills/mondlak1/camclave/skill/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/mondlak1/camclave/skill"><img src="https://agentmods.dev/badge/skills/mondlak1/camclave/skill.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00079 | $0.01093 |
| Opus 5.5 | $0.00032 | $0.00437 |
| Sonnet 5.5 | $0.00016 | $0.00219 |
| Haiku 4.5 | $0.00008 | $0.00109 |
Grade A, and why
camclave 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 2mo 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 — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.
camclave
camclave lets you see single still frames from the user's webcam. The user must have explicitly run camclave start to consent to this session; otherwise capture refuses.
There is no video capture. Every frame that lands on disk is the result of an explicit capture or scheduled snapshots write. There is no rolling buffer, no continuous recording path, no streaming. The daemon never instantiates cv2.VideoWriter and never writes any video container.
How to use it (agent-side)
Before calling anything, check the session is active:
camclave status
If no session is active, do not retry. Tell the user: "I need to see your camera. Please run camclave start in another terminal, then ask me again." That is the consent contract — the agent never instructs the user to start the session as part of an in-flow action; the user must initiate it.
To grab one frame (most common):
camclave capture --reason "read the multimeter LCD"
Always pass --reason. The reason flashes in the preview window's footer for ~4 seconds after the capture and is written to ~/.camclave/audit.jsonl. This is the user's audit trail — they should be able to glance at the screen and see exactly why each capture happened.
camclave capture (no reason) still works but logs (no reason given) and the agent SHOULD NOT use that form unless the user explicitly told it to.
Prints an absolute PNG path on stdout. Read the PNG with your image-reading tool. The preview window flashes warm cream and beeps the moment the frame is captured.
To watch over time (e.g. 3D print baby-sitting, plant logging):
camclave snapshots --every 30s --duration 10m --out /tmp/latest.png
The daemon overwrites --out on every interval. Re-read the file when you want a fresh look. Stop early with camclave snapshots-stop.
To adjust camera properties (e.g. dim breadboard, glare on an LCD, autofocus hunting on a small target):
camclave adjust --show # read current values first
camclave adjust --brightness 0.7 --exposure -5
camclave adjust --auto-focus 0 --focus 120 # disable autofocus and pin focus
What ships with it
6 files 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.
- 2mo ago First seen · 73 lines · 79 tokens per session scan A d84fbeda5594
camclave is a skill published in the GitHub repository Mondlak1/camclave (6 stars, last pushed 5mo ago), licensed MIT. It adds 79 tokens to every session and 1,093 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-08-31.
Other skills, from other repositories
agentsop-llm-engine-selection
Cross-engine decision rubric for self-hosting or recommending an LLM serving stack. Picks among vLLM, SGLang, TensorRT-LLM, TGI, llama.cpp, Ollama, and MLX as a function of (hardware × workload × constraint), not "which is fastest". Activates whenever a coder-agent must choose, defend, or migrate a serving runtime.
ros2-engineering-skills
ROS 2 engineering: rclcpp/rclpy, colcon/ament, launch, QoS/DDS, tf2/URDF, ros2control, Nav2, MoveIt 2, sensors, runtime/artifact provenance, and hardware safety. Use for development, review, debugging, and ROS 1 migration to ROS 2. Not for general C++/Python, unrelated middleware, or web/mobile tasks.
readable-verilog-generator
Use when creating, writing, reviewing, annotating, repairing, refactoring, or validating readable Verilog RTL, including synthesizable Verilog-2001 .v files, existing-RTL analysis, semantic comment annotation, testbench scaffold planning, ASIC-quality review, local or remote Vivado/xsim validation, evidence-backed…
agentic-hil
Use when a request operates this project's connected target board through the configured bench (flashing, resetting, probing, debugging, UART and CAN stimulus and feedback, firmware artifacts, test reports) or configures Agentic HIL, the safe local MCP bridge that performs them. Use it instead of invoking a debugger…
unifi-udm-pro-mcp
UniFi UDM Pro MCP Server for managing networks, connected clients, Wi-Fi settings, and device health via Model Context Protocol.
dbt-strategy
Use when creating or modifying dimensional dbt models in warehouse-backed analytics projects. Covers a four-layer warehouse architecture (sources/staging/core/marts), naming conventions, no-alias SQL rule, surrogate-key and missing-record patterns, incremental strategies, deduplication, and common project macros. Use…