Kinema's Concept Re-Search
v1.0.4Research whether a concept has been implemented and its current state. Use multi-language keywords, multi-engine cross-validation, and multi-dimensional sear...
⭐ 0· 110·0 current·0 all-time
byKinema.@leeshunee
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
The name/description (concept research) aligns with the instructions: clarify the concept, generate keywords, run broad/deep searches, fetch pages, clone repos, download PDFs, and produce a report. The requested actions are expected for a research workflow.
Instruction Scope
SKILL.md instructs the agent to use searxng-search (or ddg-search), web_fetch, and to clone GitHub repositories locally and save all artifacts under projects/research-{uuid}/. It does not instruct the agent to execute any downloaded code, nor does it attempt to read unrelated local files or environment variables. However, it assumes availability of network-fetch and git-like tooling which are not declared in the skill metadata.
Install Mechanism
No install spec or code files present (instruction-only). Nothing will be written by an installer; all actions happen at runtime via the agent's normal capabilities.
Credentials
The skill requests no environment variables, credentials, or config paths. There are no unrelated secret requests.
Persistence & Privilege
The workflow writes search results and downloaded artifacts to a project directory (projects/research-{uuid}/). This is proportional to the task, but users should be aware the skill will persist external content to the agent's filesystem.
Assessment
This skill appears coherent for concept research: it will run searches, fetch web pages, clone public GitHub repos, download PDFs, and save results to projects/research-{uuid}/. Before installing/using: ensure the agent environment provides the expected tools (web_fetch/search skills and git or repo access); be aware it will download arbitrary external content — do not execute downloaded code without inspection; store results in an isolated workspace if you worry about malicious files; consider limiting the agent's network or filesystem privileges if you want tighter control. No credentials are requested and there are no hidden install scripts according to the package contents.Like a lobster shell, security has layers — review code before you run it.
latestvk9750fjtzpdnfse916d472ph3n840pxa
License
MIT-0
Free to use, modify, and redistribute. No attribution required.