The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes a local script located at scripts/campaign_content.py with the --validate flag. This script is used to ensure derived content aligns with the verified claim ledger and project facts.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted or external source materials, including film transcripts, press materials, and campaign files, which creates a vulnerability surface.
Ingestion points: Reference files such as ../../references/CAMPAIGN_BRAND.md and various source materials (transcripts, press releases).
Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions when reading source data.
Capability inventory: The agent has the capability to write to local files (06_release/social/content_lineage.jsonl) and execute the local Python validation script.
Sanitization: There are no explicit instructions to sanitize, escape, or filter the content ingested from the external source materials before processing.

content-repurpose