Capability Radar
PassAudited by ClawScan on May 10, 2026.
Overview
This instruction-only skill transparently logs capability problems into persistent memory for later improvement, with no code, credentials, or hidden install behavior shown.
This appears safe to install as an instruction-only memory/logging skill. Before using it, make sure you are comfortable with the agent appending capability-failure notes to persistent memory and with future sessions using those notes as improvement signals.
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.
Capability-gap notes may persist across sessions and may include task context, user corrections, or operational details that future agents treat as important.
The skill creates durable memory records and instructs a future agent to read them on startup, so logged content can influence later behavior.
写入当日 `memory/YYYY-MM-DD.md` 末尾... Claude Cowork 每次会话启动时...额外扫描：`grep "🔧 from-xiaomeng: 能力雷达" memory/*.md`
Keep entries concise, avoid secrets or sensitive personal details, and periodically review or prune the memory files.
A future session may prioritize architecture-evolution work based on these markers, potentially competing with the user's immediate task.
The instruction can steer future sessions toward producing skill drafts or rule fixes when logged markers meet priority criteria.
🔴 立即处理 | 影响=高 且 重复次数≥2 | 本次会话必须产出 skill 草稿或规则修复
Treat these markers as advisory signals and confirm priority with the user before spending substantial effort or changing rules.