Agent Economy ReportAgent Economy ReportCAUTION version 1.0.0 · audited 2026-09-09
This is a documentation-only skill: three small text files that tell an agent how to use an external command-line tool called `goplaces` to query the Google Places API. There is no script, no code, and no network calls made by the skill itself. It does declare that it needs an external binary (installed via a third-party Homebrew tap, `steipete/tap/goplaces`) and a Google Places API key from the environment variable GOOGLE_PLACES_API_KEY. The documentation also mentions an optional GOOGLE_PLACES_BASE_URL variable that could redirect API traffic elsewhere, though the skill does not set it. Nothing in the files attempts to exfiltrate the API key, override agent rules, or download and run hidden code. The main things to review are trust in the third-party Homebrew tap and use of a restricted, quota-limited API key.
| Declared purpose | Provide instructions for using the `goplaces` CLI to run Google Places API (New) text search, place details, name resolution, and reviews, with human-readable or JSON output. |
| Observed behavior | The package contains only Markdown documentation and a metadata JSON file. It lists example shell commands the agent could run against the locally installed `goplaces` binary, declares an install step via Homebrew, and states that GOOGLE_PLACES_API_KEY is required. No executable code, no file writes, no direct network requests, and no persistence mechanisms are present in the package. |
| Verdict, rules only | SAFE |
| Verdict, AI | CAUTION |
| Final verdict | CAUTION (never better than either pass) |
| capability | vs purpose | detail |
|---|---|---|
| shell | needed | The skill instructs the agent to run `goplaces` CLI commands, e.g. SKILL.md:20 `goplaces search "coffee" --open-now --min-rating 4 --limit 5`. Shell execution is inherent to a CLI-wrapper skill. |
| credentials | needed | Requires the GOOGLE_PLACES_API_KEY environment variable (SKILL.md:5, 16). The key is used by the external binary; the skill files never read, copy, or transmit it. |
| install_packages | needed | Declares a Homebrew install of a third-party tap: `brew install steipete/tap/goplaces` (SKILL.md:5, 13). Code behind that tap is outside this package and unaudited. |
| network | needed | No network calls are made by the skill itself; the referenced binary contacts the Google Places API. Only documentation links to github.com and clawhub.ai appear. |
| severity | finding |
|---|---|
| medium | Installs a binary from a third-party Homebrew tap The skill asks the agent to install an executable from a personal tap (steipete/tap). The actual behavior of that binary is not contained in this package and cannot be audited here; it will receive the Google Places API key. Trust in the tap owner is required. SKILL.md:5 "install":[{"id":"brew","kind":"brew","formula":"steipete/tap/goplaces","bins":["goplaces"],"label":"Install goplaces (brew)"}] |
| low | Requires a billable API credential The skill needs a Google Places API key, which is billable. The package itself never reads or transmits the key, but an agent holding it should use a restricted key with quota/billing caps. SKILL.md:16 - `GOOGLE_PLACES_API_KEY` required. |
| low | Documented override of API endpoint The underlying binary supports redirecting all API traffic (including the API key) to an arbitrary base URL. The skill does not set this variable and no destination is hardcoded, but if the variable were set elsewhere the credential could be sent to an unintended host. SKILL.md:17 - Optional: `GOOGLE_PLACES_BASE_URL` for testing/proxying. |
| info | Publisher discloses the endpoint-override risk The skill card explicitly documents the base-URL redirection and API-key risks and recommends mitigations, which is consistent with the observed content. skill-card.md:23 Risk: GOOGLE_PLACES_BASE_URL can redirect API traffic to an unintended endpoint if set incorrectly. |
0 AI finding(s) were dropped because their file, line or quote did not match the package.
| domain | service rating | where |
|---|---|---|
| clawhub.ai | not in the service index | skill-card.md:7 |
| github.com | not in the service index | SKILL.md:4 |
| Binaries invoked | none |
| Environment variables read | none |
| Hard-coded wallets | none |
| Pipes a download to a shell | 0 |
| eval / exec / subprocess | 0 |
| base64 blobs | 0 |
| File writes | 0 |
| Persistence | 0 |
| file | lines | sha256 |
|---|---|---|
SKILL.md | 30 | 65cc8f6cc76ab9b9… |
_meta.json | 6 | ad84cebefc12e0df… |
skill-card.md | 42 | b1849b8fa6369a29… |
For agents
JSON: https://agenteconomy.report/k/goplaces.audit.json · badge: https://agenteconomy.report/k/goplaces.audited.svg ·
skill rating: /k/goplaces · commission an audit of any skill: US$ 29 per version.
The complete published package of this exact version was downloaded from the registry and read statically; nothing was executed. A deterministic pass extracts network destinations, binaries, environment variables, writes, install commands, obfuscation markers, subprocesses and wallets, each with file and line. An AI then reads the whole package with those facts and writes the summary, the capabilities and the findings under a strict schema; every finding it produces must cite a file, a line and the exact text, or it is dropped. The final verdict is the worst of the two passes. The audit does not change the skill's trust tier (policy); the author may respond through the dispute channel and the response is published here. Commissioned by: the Agent Economy Report (free program: the 150 most downloaded skills, September 2026).