Jd Writing
ReviewAudited by ClawScan on May 14, 2026.
Overview
This is mostly a coherent Chinese JD-writing skill, but it tells the agent to persistently modify the skill itself from feedback without a clear approval or review step.
Install only if you are comfortable with a JD-writing skill that may try to update its own templates. Before allowing any update, ask the agent to show the exact changes, confirm them explicitly, and separately review any referenced recruiting workflow skills.
Findings (2)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Feedback from one JD task could change how the skill behaves later, potentially affecting future users or future recruiting outputs.
This tells the agent to persistently edit the installed skill based on user feedback, but does not require explicit approval, show a diff, constrain the edit, or provide rollback. That can turn accidental or malicious feedback into future instructions.
用户指出可复用问题时，立即更新本 Skill 并在 [EVOLUTION.md](EVOLUTION.md) 记录。
Only update skill files when the user explicitly asks for a skill update, show the exact proposed diff, limit edits to the relevant template, and keep a reversible change log.
Using the full recruiting workflow may bring in additional behavior beyond this JD-writing package.
The skill may load another recruiting skill for broader workflows. That is purpose-aligned, but the referenced skill is outside the reviewed file set, so its instructions and capabilities are not assessed here.
如果任务覆盖 JD + 筛选 + 面评的完整流程，先加载 [[recruiting-skillset]]。
Review and trust the referenced recruiting-skillset before using the combined workflow.