testcase-creator
v1.0.1本技能从需求文档生成全面的测试用例文档。当用户需要从需求文档、产品规格说明或描述系统功能的文档创建测试用例时使用此技能。默认生成Markdown格式的测试用例文档；当用户明确要求生成"思维导图格式"或"xmind格式"时，会额外生成XMind思维导图文件。
⭐ 1· 159·0 current·0 all-time
bydodo@duyinghua
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 behavior: the skill generates Markdown test-case documents and optionally converts them to XMind. The included helper scripts (filename generator and md->xmind converter) are directly relevant and proportionate.
Instruction Scope
SKILL.md instructs the agent to deeply analyze input requirement documents and to always read the four included reference files before other steps — this is expected for formatting/quality enforcement. It also says that if a requirement is provided as a URL (e.g., Feishu doc), the agent should try to read it directly and, if not possible, attempt to use other skills to download it. That could lead to network access or invocation of other skills to fetch protected documents; the skill itself does not include explicit network code or credential usage.
Install Mechanism
Instruction-only install (no installation spec). Two small, readable Python scripts are included; there are no downloads from external/untrusted URLs or package installs.
Credentials
The skill declares no required environment variables, credentials, or special config paths. The scripts operate on user-provided file paths and /tmp for temporary files only.
Persistence & Privilege
Skill is not force-enabled (always:false) and doesn't request elevated persistence or modify other skills. It generates and preserves output files per its documented filename policy; this is consistent with its purpose.
Assessment
This skill looks internally consistent and the bundled scripts are readable and relevant. Before using it: (1) be mindful when supplying URLs to private documents — the skill's instructions suggest attempting to 'read' those URLs and possibly invoking other skills to fetch them, which may require credentials or expose document content; ensure you grant access only to trusted endpoints. (2) The converter writes temporary files under /tmp and saves outputs to whichever output directory you provide—verify output path permissions and that generated files don't contain sensitive secrets. (3) If you plan to let the agent autonomously fetch documents from collaboration platforms, confirm how authentication will be handled and prefer giving the document text directly when possible.Like a lobster shell, security has layers — review code before you run it.
latestvk9731dgxm9dmhqxtab0edxabjn83e7eq
License
MIT-0
Free to use, modify, and redistribute. No attribution required.