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 Darwin-Agent/Car-bench-TRACE --skill climate-air-circulationgit clone --depth 1 https://github.com/Darwin-Agent/Car-bench-TRACEWrote 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/darwin-agent/car-bench-trace/climate-air-circulation)<a href="https://agentmods.dev/skills/darwin-agent/car-bench-trace/climate-air-circulation"><img src="https://agentmods.dev/badge/skills/darwin-agent/car-bench-trace/climate-air-circulation/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/darwin-agent/car-bench-trace/climate-air-circulation"><img src="https://agentmods.dev/badge/skills/darwin-agent/car-bench-trace/climate-air-circulation.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 | $0.00084 | $0.01928 |
| Opus 5 | $0.00042 | $0.00964 |
| Sonnet 5 | $0.00017 | $0.00386 |
| Haiku 4.5 | $0.00008 | $0.00193 |
Grade A, and why
climate-air-circulation 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 9d 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 — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Set the air-circulation mode
Switch how the cabin draws air: fresh-air (outside air), recirculation, or auto. The mode may be named outright, or the user may point at "my preferred" circulation without naming it. Sometimes the control is missing. The same method handles all of these.
When this applies
"Switch to fresh-air / recirculation / auto", "the air feels stuffy, let some outside air in", "set my preferred air circulation". (Turning AC on/off is climate-air-conditioning; setting fan speed is climate-fan-speed; setting airflow direction is climate-airflow-direction.)
Tools
get_climate_settings({})— read the current circulation mode (and the rest of the climate state). Use it only to report current state; it is NOT a required precondition for setting the mode. Verify it is callable, and read its fields honestly — a field returned as"unknown"is unverifiable, not a real value.get_user_preferences({})— read the user's preferred circulation mode (the mode they like). This is what disambiguates an unnamed mode.set_air_circulation({mode})— set the circulation mode (FRESH_AIR / RECIRCULATION / AUTO); use exactly the mode named or preferred. Setting it is a direct write with no precondition and is independent of AC / window / fan state — so it does not require reading or changing those first.
Method
- Set exactly the mode named. If the user names fresh-air, recirculation, or auto, set that exact mode directly — not a "smarter" or default substitute, and without first reading climate state you don't need.
- Unnamed mode → read the preference, then set it. When the user says "my preferred" / "the one I like" instead of naming the mode, read
get_user_preferencesand set the mode it specifies. - Confirm what changed and stop. Don't loop a confirmation on a direct, unambiguous command.
Ask vs. infer (the under-specification crux)
- Unnamed circulation mode → INFER from the stored preference; do NOT ask, do NOT invent. When the user expects you to know their preferred mode, read
get_user_preferencesand set the mode it gives. Offering an open "fresh air or auto?" menu, or guessing a default, is the failure. A single confirmation of the preferred value is tolerable, but the value comes from the preference, not a guess. - Never re-ask what context already pins. If the user already named the mode, or it is fixed by the preference, act on it directly — do not ask again. Read the deciding value once, then act; don't loop on preference reads.
- Informational/options questions are NOT a call for action — answer, don't act. When the user only asks "what's the best way to improve the air?", "the air feels stagnant, what are my options?", or "what's it set to?", present the options (fresh air, auto, raise fan, etc.) and/or report the current state, then stop. Do not proactively call
set_air_circulationor any state-changing tool — wait for an explicit "do it" ("switch to fresh air", "get some air flowing"). Likewise "prepare to …" is not yet a call to act on that step. Acting on a question is over-acting. - There is no genuine open user choice here beyond the mode itself, and that is either named or fixed by preference — so asking is almost never warranted.
- In a combined request, change only what's asked plus what a tool strictly requires. When setting circulation rides along with another action (e.g. "turn on AC and set fresh air"), perform exactly the circulation change you own and let the other operation handle its own minimal preconditions. Do not pre-emptively nudge fan speed, temperature, or other actuators to a temporary value the user never requested — an unsolicited intermediate change is a failure even when the final state ends up correct, especially since a later turn may set that actuator differently.
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.
- 9d ago First seen · 63 lines · 84 tokens per session scan A dfab0981ea67
climate-air-circulation is a skill published in the GitHub repository Darwin-Agent/Car-bench-TRACE (8 stars, last pushed 6d ago), licensed MIT. It adds 84 tokens to every session and 1,928 once invoked, about $0.0004 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 2026-08-31.
Other skills, from other repositories
gke-compute-classes
Configures, optimizes, and troubleshoots GKE ComputeClasses. Use when configuring Spot VMs with on-demand fallback, targeting specific accelerators (GPUs/TPUs) or machine families, restricting ComputeClass access, or debugging pending pods related to node pool auto-creation. Do not use for cluster-level Node Auto…
jetson-diagnostic
Read-only Jetson health snapshot for identity, memory, GPU, thermal, power, storage, services, and top processes.
doca-socket-relay
Use this skill when the operator is driving the DOCA Socket Relay to bridge a socket-oriented host application onto a BlueField DPU peer without rewriting it — picking the deployment shape (in-process, sidecar, or BlueField service container), configuring the host-side socket and the DPU-side forwarding endpoint…
offensive-z-wave
Z-Wave attack methodology — sniffing with Z-Force / EZ-Wave / RTL-SDR + ZniffMobile, S0 (legacy) network-key derivation flaw and key reuse, S2 (modern) ECDH commissioning analysis, replay/injection on unauthenticated nodes, default-key brute-force on test deployments, and home-automation hub pivots. Use when targeting…
hsb-flash
Flash the FPGA on an HSB board connected to an NVIDIA devkit. Supports HSB Lattice boards (FPGA versions 2407, 2412, 2507, 2510) and Leopard Imaging VB1940 "all-in-one" cameras (FPGA versions 2507, 2510). Uses release-specific YAML manifests and board-type-specific program commands. Lattice and VB1940 commands must…
jetson-validate-image
Use after jetson-flash-image to run static BSP checks, on-target smoke/regression tests on a flashed DUT, or both. Not for build or flash steps. Triggers: validate bsp, on-target validation.