← 返回 Skills 市场
194
总下载
3
收藏
0
当前安装
6
版本数
在 OpenClaw 中安装
/install structure-prediction
功能描述
Biomolecular structure prediction tools for Chai-1, Boltz-2, and Alphafold3 via SciMiner APIs.
安全使用建议
This skill appears coherent, but consider these practical precautions before installing: 1) Verify sciminer.tech is a trusted service and you are comfortable sending sequence/MSA/template files there (these may contain sensitive research data). 2) Use a scoped API key for SciMiner (do not reuse high-privilege or long-lived credentials). 3) Confirm the '/console/api' endpoints in the SKILL.md are legitimate (the path includes 'internal', so verify the vendor docs). 4) The skill will include the share_url returned by SciMiner when summarizing results — be mindful that link recipients may be able to view full results online. 5) If you require auditability, review sciminer_registry.py (included) to ensure the listed provider/tool names and file-parameter mappings match your expectations. If any of these checks raise doubts, obtain clarification from the skill author or use alternate, vetted tools.
功能分析
Type: OpenClaw Skill
Name: structure-prediction
Version: 1.0.5
The skill provides a legitimate interface for biomolecular structure prediction using the SciMiner API (sciminer.tech). The code in `scripts/sciminer_registry.py` and the instructions in `SKILL.md` are well-documented and strictly aligned with the stated purpose of wrapping Chai-1, Boltz-2, and Alphafold3 tools. No evidence of data exfiltration, malicious execution, or harmful prompt injection was found.
能力标签
能力评估
Purpose & Capability
Name/description, SKILL.md examples, and the included sciminer_registry.py all center on invoking SciMiner tools (Chai-1, Boltz-2, Alphafold3). The single required env var (SCIMINER_API_KEY) is appropriate for a remote API integration.
Instruction Scope
Runtime instructions only describe calling SciMiner endpoints, polling for results, and uploading user-provided MSA/template files. There are no instructions to read unrelated local files or environment variables, nor to exfiltrate data to third-party endpoints beyond sciminer.tech.
Install Mechanism
No install spec and no external downloads — the skill is instruction/code-only. Nothing is written to disk by an installer as part of the skill package.
Credentials
Only SCIMINER_API_KEY is required (declared as primaryEnv). That single credential directly maps to the documented use (X-Auth-Token header) and is proportionate to the purpose.
Persistence & Privilege
always is false and the skill does not request elevated or persistent system-wide privileges. It does not attempt to modify other skills or agent configuration.
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 structure-prediction - 安装完成后，直接呼叫该 Skill 的名称或使用
/structure-prediction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
v1.0.5
- Updated documentation to specify that the SciMiner API key is free.
- Adjusted instructions to clarify users should obtain a "free SciMiner API key" from https://sciminer.tech/utility.
- No functional or code changes; documentation only.
v1.0.4
- Minor update to the skill description for improved clarity.
- No functional or interface changes.
v1.0.3
- Updated instructions to clarify that if the `SCIMINER_API_KEY` is not available, alternative tools should not be used.
- Revised API result documentation for clarity, especially the meaning of the `share_url` field.
- Minor text edits and reorganization for improved clarity and guidance.
- Removed guidance suggesting users may paste API keys directly into dialog.
v1.0.2
- Improved handling of missing API key: The agent now instructs users to obtain and provide the SciMiner API key rather than attempting alternate fallbacks.
- Clarified that the user may paste their API key into the chat input when prompted.
- Added justification for exclusively using the SciMiner API to ensure reliable, ensemble results and avoid fragmented outputs from other services.
- Updated the Notes and Prerequisites to reinforce the API key instructions and proper workflow for structure prediction tasks.
v1.0.1
- Added explicit documentation for required environment variable: SCIMINER_API_KEY.
- Updated prerequisites and notes to clarify that SCIMINER_API_KEY must be set and is used as the X-Auth-Token header.
- Introduced the "requires" and "primaryEnv" fields in the SKILL.md metadata for improved environment configuration.
- No code or file changes; documentation improvements only.
v1.0.0
Initial release of the structure-prediction skill.
- Provides access to biomolecular structure prediction tools Chai-1, Boltz-2, and Alphafold3 via SciMiner internal APIs.
- Supports multimodal complex prediction including proteins, DNA, RNA, ligands, and their interactions.
- Includes invocation patterns and example code for API use, result polling, and file uploads.
- Requires SciMiner API key and configuration instructions are included.
- Returns results with task status and shareable result URLs.
元数据
常见问题
Biomolecular Structure Prediction 是什么？
Biomolecular structure prediction tools for Chai-1, Boltz-2, and Alphafold3 via SciMiner APIs. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 194 次。
如何安装 Biomolecular Structure Prediction？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install structure-prediction」即可一键安装，无需额外配置。
Biomolecular Structure Prediction 是免费的吗？
是的，Biomolecular Structure Prediction 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Biomolecular Structure Prediction 支持哪些平台？
Biomolecular Structure Prediction 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 Biomolecular Structure Prediction？
由 SciMiner（@xiongzhp）开发并维护，当前版本 v1.0.5。
推荐 Skills