Free Search Aggregator
v1.3.0Quota-aware multi-provider web search for OpenClaw. Supports 12 search providers with automatic failover, task-level deep search (@dual/@deep), real quota ch...
⭐ 1· 539·5 current·5 all-time
byTianqu@vulcanusalex
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description, README, SKILL.md, config/providers.yaml and the Python provider implementations all describe and implement a quota‑aware multi‑provider web search with failover, discovery, health tracking and local persistence. Environment variables mentioned (BRAVE_API_KEY, EXA_API_KEY, etc.) are provider API keys appropriate to the stated providers.
Instruction Scope
Runtime instructions and the CLI (doctor, discover, search, remaining, health, gc, setup) stay within search/diagnostic scope. The discovery feature actively probes many third‑party endpoints (public SearXNG instances and several candidate search engines) and will issue network requests with the user query; results are persisted under memory/. This network probing and persistence are expected but worth noting for privacy and outbound network policy considerations.
Install Mechanism
There is no install spec (lowest install risk). The skill bundle does include Python source and a requirements.txt (requests, BeautifulSoup, PyYAML). The package assumes a Python runtime with these libraries available; dependencies are not automatically installed by the skill manifest, so the operator must ensure required packages are present. No external arbitrary downloads or URL-based installers were found.
Credentials
No mandatory credentials are declared. The code conditionally reads many provider API keys from environment variables (and treats missing keys as 'skip' behavior). The requested env vars are proportional to a multi‑provider search aggregator and are limited to provider API keys; no unrelated secrets or cloud credentials are requested.
Persistence & Privilege
The skill writes logs, health records, quota state and search artifacts under a local memory/ workspace (e.g., memory/search-cache, memory/search-index, provider‑health). It does not request always:true, does not modify other skills, and does not require system‑wide privileges. Local persistence and quota state are consistent with the described caching and health features.
Assessment
This skill appears to be what it says: a multi‑provider web search aggregator that will perform outbound HTTP requests to configured providers and public candidate search instances and store results locally under memory/. Before installing: 1) Review config/providers.yaml and disable or remove providers you don't trust (public SearXNG instances are probed by default in discovery). 2) Be mindful that queries (which may contain sensitive text) are sent to external services — do not send secrets or PII. 3) If you supply API keys, only provide keys for services you trust; keys are optional for several providers. 4) Ensure the runtime has the Python dependencies (requests, beautifulsoup4, PyYAML) since no automatic installer is specified. 5) If you are concerned about autonomous invocation or network access, restrict the skill's invocation or run it in a network‑restricted sandbox; the skill can run discovery/health probes automatically when invoked and will persist data under the workspace memory/. 6) Periodically inspect and configure retention for memory/ (gc command is available).Like a lobster shell, security has layers — review code before you run it.
latestvk972ttyey7p0tzdmqy99wr0pns82dwgz
License
MIT-0
Free to use, modify, and redistribute. No attribution required.