Life Scheduler
AdvisoryAudited by Static analysis on May 7, 2026.
Overview
No suspicious patterns detected.
Findings (0)
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.
Recent conversation themes or persona details may be carried forward into future chats through HEARTBEAT.md and history archives.
The skill intentionally converts persona and recent-chat-derived information into persistent local files that are reused in future prompt context.
“（可选）读取近期对话记录 ... 写入 HEARTBEAT.md（覆写）+ 存档到 memory/life-history/ ... 对话时 HEARTBEAT.md 自动注入”
Review or reduce reference_recent_chats, archive, and inject_to_context settings if you do not want chat-derived state persisted or reused.
The skill can continue spending model tokens and modifying context files on a schedule after setup.
The skill uses scheduled jobs that continue generating and updating state without a direct user prompt each time, though the job names and schedule are disclosed.
“Skill 会自动创建以下定时任务 ... life-scheduler-generate | 每天 07:00 ... life-scheduler-period-1 | 12:00”
Confirm the cron jobs after installation and use the documented pause, resume, or schedule-change controls if you do not want automatic updates.
Users or viewers may mistake generated fictional activities, moods, or memories for real continuity if the roleplay context is not clear.
The generation prompt asks for fictional in-character diary-like output while omitting AI/template wording, which fits roleplay but can make simulated state feel like real memory.
“用角色自己的口吻和语感写 ... 不要出现任何 AI / 系统 / 模板相关的措辞”
Use this only for agents where fictional persona continuity is intended, and keep clear disclosure if the agent interacts with people who may not understand it is simulated.