InfoSeek
v2.0.0Deep web information search and archival skill for comprehensive research on persons, organizations, or products. Uses multiple search engines (Baidu, Tavily...
⭐ 0· 56·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description (deep web search + archival) align with the included helper script (URL normalization, SQLite deduplication, file storage) and the declared requirement of python3 and OPENCLAW_WORKSPACE. The script explicitly handles local file and DB operations and does not perform network searches itself, which fits the model where the agent or other 'search' skills perform crawling.
Instruction Scope
SKILL.md instructs the agent to use external search/browser skills to fetch pages and to run the local helper script for normalization, deduplication, and saving. It does not instruct the agent to read arbitrary unrelated files or extra environment variables beyond OPENCLAW_WORKSPACE. Minor issues: inconsistent naming for required skills (e.g., 'tavily' vs 'tavily-search', 'Multi-Search-Engine' vs 'multi-search-engine') and a reliance on other skills being present in workspace/skills; these appear to be sloppy bookkeeping rather than malicious scope creep. Also, the workflow encourages high-volume scraping (e.g., '100+ pages' on Baidu) — a functional concern (rate limits, TOS, IP blocking, legal/ethical risk), not a code/credential mismatch.
Install Mechanism
No install spec is provided (instruction-only skill with one helper script included). That is low-risk: nothing is downloaded from remote URLs and the script will only be written to the agent environment when this skill is installed. The helper script is plain Python, readable, and contains no obfuscated code or hidden remote endpoints.
Credentials
The only declared primary credential is OPENCLAW_WORKSPACE (a workspace path used to store archives and check for other skills). No API keys or unrelated secrets are requested. The workspace access is necessary and proportionate for saving archives and database files.
Persistence & Privilege
always is false (no forced always-on presence). The skill writes files and an SQLite DB under the workspace (expected for an archival tool) but does not request elevated system-wide configuration changes or access to other skills' configs.
Assessment
This skill appears to do what it says: it expects a workspace path and a readable/writable folder to store archives and an included Python helper script to manage deduplication and file storage. Before installing, consider: 1) Trust/source — the package has no homepage and an unknown source; review the full helper script yourself (it is included) and confirm you trust the publisher. 2) Dependencies — the skill expects other search/browser skills to exist in {workspace}/skills; ensure those are genuine and named exactly as SKILL.md expects (there are some naming mismatches in the instructions). 3) Legal & operational risk — the workflow encourages high-volume crawling (e.g., 100+ pages); ensure you comply with target sites' terms of service, robots.txt, and avoid overloading sites. 4) Workspace safety — the skill will create infoseek-archives/ and an SQLite DB under OPENCLAW_WORKSPACE; point OPENCLAW_WORKSPACE to an isolated location if you don't want data mixed with other agent state. 5) Rate limiting & secrets — the helper script does not exfiltrate data or call remote endpoints, but other search/browser skills might. Verify those dependent skills before use. If you want higher assurance, ask the publisher for a homepage or repository, or run the skill in a sandboxed workspace first.Like a lobster shell, security has layers — review code before you run it.
latestvk978zzsxs5evanzq51fbznr67n84dmet
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
Binspython3
Primary envOPENCLAW_WORKSPACE