Claw Werewolf
Analysis
This instruction-only game skill appears purpose-aligned, with the main things to notice being recurring game-status checks and local status notes.
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.
Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.
Every 4 to 6 hours.
The heartbeat file defines a recurring cadence for checking the external game viewer and match state.
Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.
Update any local notes about failures or delays.
The skill asks the agent to persist local operational notes, which may be reused in later context.