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 place-phone-callgit 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/place-phone-call)<a href="https://agentmods.dev/skills/darwin-agent/car-bench-trace/place-phone-call"><img src="https://agentmods.dev/badge/skills/darwin-agent/car-bench-trace/place-phone-call/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/place-phone-call"><img src="https://agentmods.dev/badge/skills/darwin-agent/car-bench-trace/place-phone-call.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 | $0.00135 | $0.02210 |
| Opus 5 | $0.00068 | $0.01105 |
| Sonnet 5 | $0.00027 | $0.00442 |
| Haiku 4.5 | $0.00014 | $0.00221 |
Grade A, and why
place-phone-call 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 10d 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 — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Place a phone call
The user wants a number dialed — a restaurant or hotel found earlier, a charging station to reserve a plug, a contact. This skill does exactly one thing: pass a real phone number to the call tool. The number must already exist in the context — a found POI's own result, a station's reservation number, a resolved contact's detail — it is never invented. Finding the POI is poi-search; resolving a contact's number is contacts-lookup.
When this applies
"Call it", "ring them", "phone the one I picked", "reserve the plug", "call the station." A number (or a source that yields one) is already in the conversation: the POI the user chose, the station selected for a reservation, or a contact's looked-up detail.
Tools
call_phone_by_number({phone_number})— dial. The argument must be the number from the specific source the context provides: the chosen POI's own search result, or the found station's number. Before relying on it, confirm it is actually callable; if it isn't available to you, no call can be placed and you must handle that honestly (see below).
Method
- Take the number from its real source. Use the phone number that belongs to the specific POI the user picked, or the station being reserved — exactly that listing's number, not a different one, and never a guessed or made-up number.
- Place the call last. Placing a call ends the conversation automatically, so dial only once every other intent in the turn is already done. In a trip flow this means navigation must be set first (that set is the
navigation-startskill's job; this skill places the call afterward) and any conditional check the user attached to the call (e.g. "call only if the time is under N minutes") must already be evaluated and satisfied. Don't dial before the route is committed or before a stated condition is met. - Dial with a clean, well-formed tool call.
call_phone_by_number({phone_number:<the source's own number>}). Emit the call as an actual tool call, not as narration text that merely describes calling. Do not pair the call with a spoken user message in the same turn: per policy, sending a message together with a tool call causes one of them to be dropped, and the call itself ends the conversation anyway, so a co-emitted "Calling them now" confirmation is both unnecessary and risks the call being lost. If you want a brief lead-in, keep it to its own prior turn; when it's time to dial, just issue the call. - Read the result before confirming. Only say the call was placed if the tool result actually reports it (e.g. the call-placed field is true). If that field comes back as
"unknown"or the call doesn't return a clear success, treat it as unverified — say you've attempted it but can't confirm it connected, and surface the number; don't assert "done" or "reservation made" on faith. Then stop.
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.
- 10d ago First seen · 66 lines · 135 tokens per session scan A 0a4ac3144cfc
place-phone-call is a skill published in the GitHub repository Darwin-Agent/Car-bench-TRACE (8 stars, last pushed 7d ago), licensed MIT. It adds 135 tokens to every session and 2,210 once invoked, about $0.0007 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
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…
next-partial-prefetching-adoption
Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…