The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted subcontractor bid documents (PDFs, emails, spreadsheets), which constitutes an indirect prompt injection surface.
Ingestion points: External bid documents as described in the workflow section of SKILL.md.
Boundary markers: SKILL.md instructs the agent to treat documents as data only and ignore any embedded instructions.
Capability inventory: The skill uses a custom Python script for data processing with no network access or high-privilege operations.
Sanitization: Input data is strictly validated against a JSON schema by scripts/level_bids.py before processing.

bid-leveling