← 返回 Skills 市场
outdog-hwh

targeted-tech-research

作者 outdog-hwh · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ 安全检测通过
58
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install targeted-tech-research
功能描述
Performs in-depth, vendor-specific technical research on solutions/products with standardized, fact-verified breakdowns of hardware, software, co-design, and...
安全使用建议
This skill appears internally consistent and appropriate for public-source technical research. Before installing: (1) confirm you trust the platform sub-skills it will call (web-scraper, playwright-scraper) because those perform network IO, (2) avoid asking it to fetch private or behind-login material (the Skill states it will abandon paywalled pages), and (3) verify any high-impact or safety-sensitive technical derivations against original sources (the Skill will label derived content but you should cross-check). If you do not want the agent to run research autonomously, restrict the skill's invocation in your agent settings.
功能分析
Type: OpenClaw Skill Name: targeted-tech-research Version: 1.0.0 The skill bundle is designed for automated technical research and product architecture breakdown. It utilizes Python scripts for safe text processing, such as stripping HTML tags (scripts/compress_content.py) and generating report narratives from local metadata (scripts/generate_narrative.py). The instructions in SKILL.md and references/prompts.md are strictly aligned with the stated purpose, emphasizing source attribution and technical accuracy without any evidence of data exfiltration, malicious execution, or harmful prompt injection.
能力评估
Purpose & Capability
Name/description match the implementation: SKILL.md defines a research workflow and explicitly delegates network scraping to sub-skills. Included helper scripts (content compression, narrative generation) and a pre-flight input validator are appropriate for a research orchestration skill.
Instruction Scope
Runtime instructions constrain actions to orchestrating research, enforcing source/derivation annotations, and delegating scraping. SKILL.md explicitly prohibits accessing paywalled/login-protected pages and claims not to perform direct network requests. No instructions request arbitrary system files, secrets, or unrelated data.
Install Mechanism
No install spec is provided (instruction-only with small local helper scripts). No downloads from external URLs or package installs are present in the manifest.
Credentials
The skill declares no required environment variables, credentials, or config paths. All included scripts operate on local input/output files and metadata; there are no secret-looking env var references.
Persistence & Privilege
Skill is not marked always:true and uses normal autonomous invocation defaults. It does not request system-wide configuration changes or persistent elevated privileges.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install targeted-tech-research
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /targeted-tech-research 触发
  4. 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
Targeted deep technical research Skill initial release: - Provides structured, in-depth research on vendor-specific technical solutions/products, covering Hardware Breakdown, Software Breakdown, Hardware-Software Co-Design, and Technical Benchmarking. - Strictly distinguishes between publicly verifiable facts and technical derivations, ensuring clear source attribution in all outputs. - Implements an interactive onboarding questionnaire for research precision, priority, granularity, terminology explanation, and source preferences. - Delegates all scraping tasks to web-scraper or playwright-scraper based on page type and dynamic site rules, with robust failure and degradation handling. - Features a five-step workflow, starting with a reconnaissance phase and user confirmation before proceeding to deep-dive technical analysis. - Prohibits generalized industry knowledge and marketing language, focusing solely on objective, structured technical breakdowns.
元数据
Slug targeted-tech-research
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

targeted-tech-research 是什么?

Performs in-depth, vendor-specific technical research on solutions/products with standardized, fact-verified breakdowns of hardware, software, co-design, and... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 58 次。

如何安装 targeted-tech-research?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install targeted-tech-research」即可一键安装,无需额外配置。

targeted-tech-research 是免费的吗?

是的,targeted-tech-research 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

targeted-tech-research 支持哪些平台?

targeted-tech-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 targeted-tech-research?

由 outdog-hwh(@outdog-hwh)开发并维护,当前版本 v1.0.0。

💬 留言讨论