skill-optimizer
PassAudited by ClawScan on May 1, 2026.
Overview
This instruction-only skill is coherent with its stated purpose, but users should notice that it may read conversation logs and, after confirmation, edit another skill's SKILL.md file.
This skill appears safe to install as an instruction-only optimizer. Before using it, sanitize conversation logs, confirm exactly which skill file will be edited, review every proposed change, and keep a backup of the original SKILL.md.
Findings (3)
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.
If approved, the skill can make persistent changes to another skill's instructions.
The skill instructs the agent to modify another skill's instruction file after user confirmation. This is aligned with optimization work, but editing skill files can change future agent behavior.
用户确认后，直接修改目标技能的 SKILL.md
Review the proposed edits carefully, keep a copy of the original SKILL.md, and approve only the specific changes you want applied.
Private information in conversation logs may be brought into the agent's working context during analysis.
The skill is designed to ingest conversation logs, either pasted by the user or read from a file path. This is necessary for its purpose, but logs can contain sensitive data or untrusted prompt-like text.
获取对话内容：用户直接粘贴文本，或提供文件路径（读取文件）
Remove secrets, credentials, personal data, and unrelated private content from logs before using the skill.
A poor or overbroad edit could affect future sessions where the modified skill is invoked.
The workflow expects changes to be written to the target skill and then summarized. Persistent skill-file changes may influence future uses of that skill.
修改完成后展示变更摘要
Apply changes in small batches, test the modified skill, and retain a rollback copy.