deepresearchpro
v1.0.0专注跨语言、多平台深度检索，产出含多维视角和严格引用的可追溯调研报告。
⭐ 0· 122·1 current·1 all-time
byJason Abbott@aptjason
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 research across media and academic sources) align with the provided SKILL.md and the included helper script (query generation, parsing, citation formatting). There are no unrelated environment variables, binaries, or config paths requested.
Instruction Scope
SKILL.md explicitly instructs the agent to use web_fetch/browser tools to gather information from news sites, social media, and academic databases and to produce cited reports. It does not instruct reading local files, env vars, or sending data to unknown external endpoints. It does instruct immediate execution upon receiving a topic, which is consistent with an agent designed to run searches on demand.
Install Mechanism
No install spec is provided (instruction-only). A local Python helper (scripts/research_tools.py) and markdown templates are included but there is no download-from-URL or package installation step. This minimizes install-time risk.
Credentials
The skill declares no required environment variables or credentials. Although the instructions reference paywalled platforms (e.g., CNKI) that might need credentials in practice, the skill does not request any secrets or unrelated credentials.
Persistence & Privilege
always is false; the skill is user-invocable and allows autonomous invocation (platform default). The skill does not request system-wide persistence or modify other skills' configs.
Assessment
This skill is internally consistent for doing web-based research: it will perform many live web fetches (news sites, social media, Google Scholar) and assemble a cited report. Before installing/using: be aware it requires network access and may encounter paywalled or login-protected pages (you would need to supply credentials if you want those pages accessed — do not share secrets unless you explicitly trust and control how they are used). Expect the agent to fetch many public pages (possible rate limits or terms-of-service considerations for some sites). If you need stricter control, run it in a sandboxed agent, limit which domains it may query, or instruct it to avoid scraping authenticated/paywalled content. If your research involves sensitive or private information, avoid providing that data to the agent.Like a lobster shell, security has layers — review code before you run it.
latestvk9741d0z4yfj93fdz0rjjafnh9833phf
License
MIT-0
Free to use, modify, and redistribute. No attribution required.