searchwire-skillSAFE
Agent skill: real keyword search volume, Google results and product-name checks for Claude Code, Codex and Cursor, through the Searchwire API
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
Gives a coding agent (Claude Code, Codex, Cursor and others) real SEO numbers instead of guesses:
- Keyword search volume, ad price per click and competition, for up to 1,000 keywords in one call
- Google results for a keyword: who ranks, with URL and title
- Name checks: whether a product name is free as a domain, on GitHub and on npm
It is one SKILL.md that teaches the agent to call the Searchwire HTTP API. There is nothing to run and no MCP server.
Install
With the skills CLI:
npx skills add LambdaLabsHQ/searchwire-skill
Or by hand, for Claude Code:
mkdir -p .claude/skills/searchwire curl -s https://searchwire.dev/skill.md -o .claude/skills/searchwire/SKILL.md
Other agents can read the full API docs at .
Try it
Domain and name checks work without a key:
Is "pagenta" free as a .com, .dev and .ai domain, on GitHub and on npm?
Search volume and Google results need a key:
List 200 keywords our docs could rank for, get their US search volume from Searchwire in one call, keep the 20 best, check who ranks for the top 5, and write a content plan.
That run costs about $0.07.
What an answer looks like
Asked on 10 October 2026, US:
Key and price
A key comes with prepaid credit, from $10. There is no sign-up and no subscription, credit does not expire, and unused credit is refunded on request within 30 days. Buy one at , or let the agent start the purchase: it gets a payment link for you and picks up the key once you have paid.
c95d6a04c798OBSERVED · 2026-10-10Install
Commands as the repository documents them. They are shown, not run.
npx skills add LambdaLabsHQ/searchwire-skill
Host compatibility
What the documentation claims. We have not run a compatibility test.
| Host | Status | Notes |
|---|---|---|
| claude-code | mentioned | |
| codex | mentioned | |
| cursor | mentioned |
What it tells the agent
The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.
---
name: searchwire
description: Look up SEO data for a task - keyword search volume, Google results for a keyword, whether domains are registered, and whether a product name is free (domains, GitHub, npm). Use when choosing keywords, sizing demand, checking competitors in Google, or naming a product.
---
# Searchwire
Base URL: https://searchwire.dev. Full docs: https://searchwire.dev/llms.txt
Calls are POST with a JSON body. Send `Authorization: Bearer $SEARCHWIRE_KEY` when you have a key.
Without a key, domain and name checks and cached data work for 20 calls a day, 10 keywords a call.
- Search volume (batch up to 1,000 keywords in one call, it is cheaper):
`curl -s https://searchwire.dev/v1/keywords/volume -H "Authorization: Bearer $SEARCHWIRE_KEY" -H "content-type: application/json" -d '{"keywords":["ai wiki","living docs"],"country":"us"}'`
- Google results: `{"keyword":"ai wiki","country":"us","depth":10}` to https://searchwire.dev/v1/serp
- Domains: `{"domains":["example.dev","example.com"]}` to https://searchwire.dev/v1/domains/check
- Product name: `{"name":"example","tlds":["com","dev","ai"]}` to https://searchwire.dev/v1/names/check
Work in this order: collect candidate keywords, get volume for all of them in one call,
narrow down, then check Google results for the few that matter (mind `observed_at`).
Prices: search volume $0.02 a call plus $0.0002 a keyword; Google results $0.002 collected, $0.005 live
(`"source":"live"`); domain, GitHub and npm checks free. Each key has a daily budget (`POST /v1/account`).
Every response has `cost` (USD) and each item has `source`, `origin`, `fetched_at`, `cached`.
Items that could not be answered carry `error`; do not invent values for them.
No key yet: `POST https://searchwire.dev/v1/billing/checkout` with `{"amount_usd":10}` returns `checkout_url` (ask a person to pay)
and `claim_url` (GET it after payment to receive the key). Save the key where it outlives this session, for
example as `SEARCHWIRE_KEY` in a secret store or an untracked `.env`, and tell the person where it is.
A person can see the key, the balance and the spend at https://searchwire.dev/account by signing in with the email address used at checkout.Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | NA |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- none-observed
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (0)
No findings outside the package's declared scope.
Gates applied: no_behavioural_pass.
c95d6a04c798full audit observations/trust-audit/skill/lambdalabshq__searchwire-skill.json · Report an issue / request a re-scanAudit history
Every audit this skill has had.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-10 | c95d6a04c798 | SAFE | B | 89 | first audit |
Questions
What does the searchwire-skill skill do?
Agent skill: real keyword search volume, Google results and product-name checks for Claude Code, Codex and Cursor, through the Searchwire API
Is searchwire-skill safe to install?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.
What can searchwire-skill access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
Which assistants does searchwire-skill work with?
Its documentation mentions claude-code, codex and cursor. That is what the text claims, not a compatibility test we ran.
How current is this page?
The grade is for one exact copy of the source (c95d6a04c798), read on 2026-10-10. The repository is watched, and a new audit runs when it changes — this is the first audit.