The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes instructions in SKILL.md for users to execute scripts/selftest.py for installation and maintenance. This script locally validates data structures using asset_check.py without network access or privileged operations.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted drama scripts. It addresses this risk through strict instructional constraints (e.g., rules AST-01 through AST-15) that require the agent to rely solely on mechanical facts and visible evidence rather than guessing or following potentially malicious instructions embedded in the script data.

short-drama-assets