← 返回 Skills 市场
1163
总下载
0
收藏
1
当前安装
4
版本数
在 OpenClaw 中安装
/install unifuncs-deep-research
功能描述
Use UniFuncs Deep Research API to run in-depth research and generate long-form reports (10,000 words or more). Use this skill when users request deep researc...
安全使用建议
This skill appears to implement a legitimate UniFuncs API client for producing long research reports, but note these points before installing: (1) the code and SKILL.md require UNIFUNCS_API_KEY, but the registry metadata incorrectly lists no required env vars — supply only an API key you trust and expect to be used by this skill; (2) the scripts create/read stream files (temp or --stream-file) and will write to any writable path you supply — avoid pointing --stream-file at sensitive locations; (3) deep-research-report.py imports subprocess (the rest of that logic is truncated here) — review the full script to confirm it doesn't execute arbitrary shell commands with user-controlled input; (4) the skill contacts api.unifuncs.com and will send your query and options to that external service — do not send secrets or sensitive data to the skill; (5) the source/homepage are missing, so you should only install if you trust the provider or can audit the full scripts. If you want higher assurance, ask the publisher for a canonical homepage/repo and a clear update to the registry metadata to declare UNIFUNCS_API_KEY as a required credential, and request the full deep-research-report.py content be reviewed for subprocess usage.
功能分析
Type: OpenClaw Skill
Name: unifuncs-deep-research
Version: 0.0.7
The skill bundle provides a legitimate interface for the UniFuncs Deep Research API, allowing for long-form report generation. The Python scripts (deep-research-report.py, deep-research-create-task.py, and deep-research-query-task.py) use standard libraries for API communication and include a background worker mechanism to handle long-running tasks. The SKILL.md instructions specifically include a safety-oriented 'Mandatory second confirmation' step to prevent accidental high-cost or long-duration executions, and no evidence of data exfiltration, obfuscation, or malicious intent was found.
能力评估
Purpose & Capability
Name/description match the included code: the three Python clients call api.unifuncs.com endpoints (create_task, chat/completions, query_task) and implement options for streaming, output length, domain allowlist/blacklist, etc. These requirements are consistent with a deep-research/reporting tool.
Instruction Scope
SKILL.md restricts runtime to running the provided Python scripts (allowed-tools: Bash(python3:*)) and enforces a mandatory second confirmation step before invocation, which is a positive safety measure. The scripts read an API key from UNIFUNCS_API_KEY, perform network requests to api.unifuncs.com, create/read temporary or user-specified stream files, and may spawn subprocesses (deep-research-report.py imports subprocess). The scripts do not appear to access other environment variables or unrelated system configuration, but they do write to arbitrary writable paths if the user supplies a --stream-file path; that could overwrite files if misused. The use of subprocess is present in the code base (truncated portion not visible) — this should be inspected to ensure it doesn't execute arbitrary uncontrolled commands.
Install Mechanism
No install spec (instruction-only with bundled scripts). Nothing is fetched from remote URLs during install; the risk surface is limited to executing the included Python scripts. This is lower-risk than arbitrary remote downloads, but executing bundled code still requires trust in the source.
Credentials
The SKILL.md and all three Python files require an API key via the environment variable UNIFUNCS_API_KEY. However, the registry metadata lists 'Required env vars: none' and 'Primary credential: none' — this is an inconsistency. Requesting a single service API key is reasonable for this skill's purpose, but the metadata omission is a coherence problem and could mislead users about required credentials.
Persistence & Privilege
The skill is not always:true and does not claim to persistently modify system or other skills' configuration. It writes temporary/stream files as part of streaming behavior, which is expected for long-running streaming outputs. No privileged system modifications are present in the visible code.
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 unifuncs-deep-research - 安装完成后，直接呼叫该 Skill 的名称或使用
/unifuncs-deep-research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
v0.0.7
Enforces explicit second-step user confirmation before starting deep research.
- Added a mandatory process: summarize the research request and require clear user confirmation before running any Deep Research scripts.
- Users must confirm the topic, acknowledge runtime and cost, and review any special options before the skill proceeds.
- Repeats confirmation step if user changes research topic/options.
- Previous usage guidelines updated to reflect this two-step confirmation requirement.
v0.0.5
No user-visible changes in this version.
- No file changes detected; documentation and functionality remain the same.
v0.0.3
unifuncs-deep-research 0.0.3
- Major refactor: transitioned from Bash API scripts to independent Python client scripts for deep research.
- Added three entry scripts: deep-research-report.py (sync), deep-research-create-task.py (async create), deep-research-query-task.py (async query).
- Upgraded configuration and usage instructions for clearer setup and workflow.
- Expanded and clarified output format options (report, summary, wechat-article, etc.).
- Removed obsolete Bash script and reference documentation files.
v0.0.1
unifuncs-deep-research 1.0.0 released
- Initial release with UniFuncs API integration for deep research.
- Generates comprehensive, multi-format reports (report, summary, WeChat, Xiaohongshu, Zhihu).
- Customizable parameters for research theme, model selection, and output type.
- Simple configuration using UniFuncs API Key.
- Includes clear usage instructions and parameter details.
元数据
常见问题
Unifuncs Deep Research 是什么？
Use UniFuncs Deep Research API to run in-depth research and generate long-form reports (10,000 words or more). Use this skill when users request deep researc... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 1163 次。
如何安装 Unifuncs Deep Research？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install unifuncs-deep-research」即可一键安装，无需额外配置。
Unifuncs Deep Research 是免费的吗？
是的，Unifuncs Deep Research 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Unifuncs Deep Research 支持哪些平台？
Unifuncs Deep Research 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 Unifuncs Deep Research？
由 UniFuncs（@vinlic）开发并维护，当前版本 v0.0.7。
推荐 Skills