Screen recommendation loop
PassAudited by ClawScan on May 1, 2026.
Overview
This instruction-only skill is coherent for movie/anime recommendations, with only purpose-aligned notes about storing preference history and sending scheduled follow-ups.
Before installing, decide whether you want the agent to keep a persistent media-preference history and send follow-up reminders. If you use it, keep the stored record free of personal identifiers and make sure you can pause, reset, or delete the loop.
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.
Your viewing history and preference notes may be retained for future recommendations.
The skill explicitly stores viewing status, tags, and feedback in persistent local state, which is expected for long-term recommendation personalization but still creates a taste/profile record.
Keep per-title state: ... status ... tags (optional) ... note (optional, one-line user feedback). This can live in JSON or SQLite.
Use a clearly named local store, avoid identifiers, and provide an easy way to review, reset, or delete the recommendation history.
The agent may remind you later and continue the recommendation cycle unless you stop or pause it.
The skill instructs the agent to continue a recommendation loop and initiate later follow-ups, which is disclosed and purpose-aligned but represents ongoing autonomous behavior.
Schedule follow-up based on title type. ... Movie recommendation: follow up at recommendedAt + 7 days ... Anime/series recommendation: follow up at recommendedAt + 30 days
Confirm opt-in before starting the loop, make the next follow-up time visible, and support simple pause/stop commands.