← 返回 Skills 市场
88
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install target-novelty-scorer-1
功能描述
Score the novelty of biological targets through literature mining and.
安全使用建议
This skill is functionally inconsistent: its documentation promises real PubMed/PMC retrieval and lists NCBI/API dependencies, but the included script simulates results locally and doesn't use network calls or credentials. Before installing or using it: (1) Do not provide API keys to this skill — the registry metadata does not require them and the code doesn't use them. (2) Inspect scripts/main.py fully and run python -m py_compile scripts/main.py in a safe environment to confirm behavior. (3) If you need real literature mining, ask the author or maintainer to: implement actual PubMed/E-utilities calls (or clarify it's intentionally a demo), update requirements.txt, and declare any required environment variables explicitly. (4) Treat outputs as synthetic/demo data until the code is confirmed to perform real retrieval; do not use this for sensitive research decisions without verification.
功能分析
Type: OpenClaw Skill
Name: target-novelty-scorer-1
Version: 1.0.0
The skill bundle is a legitimate, albeit early-stage, tool for scoring biological target novelty. The primary script, `scripts/main.py`, currently uses a simulated implementation for literature searching via `numpy.random` and contains no evidence of malicious execution, data exfiltration, or unauthorized network activity. While `SKILL.md` mentions dependencies like `biopython` and `requests` that are missing from `requirements.txt`, this appears to be a development oversight rather than a security risk.
能力评估
Purpose & Capability
SKILL.md and description state the tool retrieves literature (PubMed/PMC) and lists NCBI API key, requests, biopython, and pandas as dependencies. The shipped code (scripts/main.py) does not perform network calls; PubMedSearcher.search() returns simulated/randomized data. Declared capabilities (real literature retrieval) do not match the actual implementation (local simulation).
Instruction Scope
Runtime instructions are limited to running the packaged script and validating inputs; they do not ask the agent to read unrelated files or export secrets. However SKILL.md instructs users to provide an NCBI API key and mentions editing an in-file CONFIG block — the code does not require or use such credentials, creating a gap between instructions and actual behavior that could confuse users or cause them to expose keys unnecessarily.
Install Mechanism
No install spec — skill is instruction-only with a bundled script. That minimizes installation risk. Minor mismatch: SKILL.md lists many Python dependencies, but requirements.txt only lists dataclasses and numpy; this is an implementation/documentation inconsistency rather than a direct install risk.
Credentials
SKILL.md lists 'NCBI API Key' (and optional Europe PMC) as API requirements, but the registry metadata declares no required environment variables and the included script does not use any environment credentials. Asking for an API key in docs while not declaring or using it is disproportionate and could lead users to provide secrets unnecessarily if they attempted to wire them in.
Persistence & Privilege
The skill does not request persistent or elevated privileges (always: false). It does not install services or modify other skills. Autonomous invocation is allowed by default but is not combined with broad credential access or system modifications.
如何使用
- 确保已安装 OpenClaw(本地或 Docker 部署)
- 在对话框中输入安装命令:
/install target-novelty-scorer-1 - 安装完成后,直接呼叫该 Skill 的名称或使用
/target-novelty-scorer-1触发 - 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
Initial release of Target Novelty Scorer.
- Scores the novelty of biological targets based on literature mining from academic databases (e.g., PubMed, PMC).
- Features include literature retrieval, multi-dimensional novelty scoring (0–100), trend analysis, cross-validation, and detailed report generation.
- Provides both basic and advanced command-line usage, supporting output in text, JSON, or CSV.
- Assesses targets using five main scoring criteria: research heat, uniqueness, research depth, collaboration network, and temporal trends.
- Includes security, risk, and evaluation checklists to ensure safe and reliable operation.
- Requires Python 3.9+, with dependencies including requests, pandas, biopython, and numpy.
元数据
常见问题
Target Novelty Scorer 是什么?
Score the novelty of biological targets through literature mining and. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 88 次。
如何安装 Target Novelty Scorer?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install target-novelty-scorer-1」即可一键安装,无需额外配置。
Target Novelty Scorer 是免费的吗?
是的,Target Novelty Scorer 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
Target Novelty Scorer 支持哪些平台?
Target Novelty Scorer 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。
谁开发了 Target Novelty Scorer?
由 AIpoch(@aipoch-ai)开发并维护,当前版本 v1.0.0。
推荐 Skills