The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill performs several automated downloads to bootstrap local game databases.
scripts/fetch_fgo_db.py downloads Fate/Grand Order data from api.atlasacademy.io.
scripts/fetch_ygo_db.py downloads Yu-Gi-Oh! card information from db.ygoprodeck.com.
scripts/fetch_pokemon_db.py fetches species data from GitHub repositories.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from untrusted external sources, creating a surface for indirect prompt injection.
Ingestion points: The script scripts/fetch_story_ref.py retrieves search result snippets from DuckDuckGo, and scripts/align_ref_timeline.py parses text from external SRT subtitle files and JSON chunks.
Boundary markers: Absent. The skill does not implement delimiters or specific instructions to ensure the agent ignores natural language commands that might be embedded in the fetched synopses or subtitles.
Capability inventory: The skill is designed to execute shell commands (yt-dlp, local Python scripts) and perform network requests via urllib.
Sanitization: Absent. The skill extracts names and text directly from external sources for mapping without validation or filtering for prompt injection markers.
[COMMAND_EXECUTION]: The SKILL.md instructions explicitly guide the agent to execute shell commands, including yt-dlp for subtitle extraction and several Python scripts for database management and timeline alignment.

anibon-world-identity