疾病最新药物研究报告
v1.0.0A disease-to-innovative-drug analysis skill for biomedical question answering. It is used to answer questions such as "What innovative, cutting-edge, pipelin...
⭐ 0· 111·0 current·0 all-time
by@inbreak
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description match the bundled code and instructions. Required binaries (bash, python3), the requests and langchain-tavily packages, and the single env var TAVILY_API_KEY map to the declared functionality (local ChEMBL and ClinicalTrials adapters plus an optional Tavily-backed Search). No unrelated credentials or tools are requested.
Instruction Scope
SKILL.md instructs the agent to use only the included local tools (local_tools/*.py and run_tool.sh) and to avoid calling arbitrary external search except as an explicit, permissioned fallback. The runtime instructions do not request reading unrelated system files or secrets beyond TAVILY_API_KEY. The requirement to match the report template is explicit and narrow.
Install Mechanism
This is instruction-first and has no high-risk download/install steps. Dependencies are standard Python packages (requests, langchain-tavily) declared in SKILL.md and requirements.txt — reasonable for the stated function but moderate risk relative to pure instruction-only skills because pip installs run arbitrary code from PyPI.
Credentials
Only TAVILY_API_KEY is required and is specifically tied to the Search adapter. No unrelated secret variables or system config paths are requested. The skill does network calls to public APIs (ChEMBL, ClinicalTrials.gov) which is expected for this use case.
Persistence & Privilege
always is false and the skill is user-invocable. It does not request installation that modifies other skills or system-wide settings. The skill can be invoked autonomously by the agent (disable-model-invocation is false), which is platform default and not by itself a disqualifier.
Assessment
This skill appears coherent, but review these points before installing: (1) It will make outbound requests to public APIs (ChEMBL, ClinicalTrials.gov) and — if used — to Tavily via the langchain-tavily package, so any queries you run may be transmitted to third parties. Avoid sending sensitive patient data into queries. (2) You must provide a TAVILY_API_KEY to enable the Search fallback; protect that key like other secrets. (3) The code depends on pip packages (requests, langchain-tavily); installing them executes third-party package code from PyPI — review versions and install in an isolated environment. (4) The repository is marked Proprietary; check licensing and usage restrictions if you plan to redistribute or use in production. (5) If you want stronger guarantees, audit the included local_tools code (it's small and readable) and run in a restricted network environment or without the Tavily key to avoid external search traffic.Like a lobster shell, security has layers — review code before you run it.
latestvk974kh0pnrc0xtzed80hqqmwe1837at0
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
💉 Clawdis
OSmacOS · Linux
Binsbash, python3
EnvTAVILY_API_KEY