super-search-engine

Multi search engine integration with 17 engines (8 CN + 9 Global). Supports advanced search operators, time filters, site search, privacy engines, and Wolfra...

This skill appears coherent and low-risk: it only builds search URLs and shows web_fetch examples and asks for no keys or installs. Things to consider before enabling: (1) the agent will perform network requests to third-party search engines — review your organization's network/egress policy and privacy requirements; (2) the documented advanced operators can be used to find sensitive information on the web (e.g., filetype:txt, intext:password) — avoid running broad discovery queries against private domains; (3) the publisher/source is unknown and there is no homepage — exercise normal caution (sandbox or limit autonomous use) if you plan to let the agent run queries without supervision. The stray text in the SKILL.md looks like a harmless typo but indicates limited editorial review.

Type: OpenClaw Skill Name: super-search-engine Version: 1.0.0 The skill bundle provides a comprehensive search engine integration but includes highly risky 'Google Dorking' templates in `references/international-search.md`, such as queries specifically designed to find passwords in text files (`intext:password filetype:txt`) and administrative login pages (`inurl:login admin`). Additionally, the `SKILL.md` file contains a nonsensical string ('brown anyway tonight...') in its description, which is a common pattern in prompt injection or canary testing. While the skill primarily functions as a search aggregator, these elements facilitate unauthorized reconnaissance and suggest potential subversion attempts.