The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of data from external sources, creating a surface for indirect prompt injection if the data is subsequently processed by an LLM without isolation.
Ingestion points: FileExtractor.extract (local files), extract_paginated (external APIs), and extract_from_db (databases) defined in SKILL.md.
Boundary markers: The skill does not implement specific delimiters or 'ignore' instructions to prevent the agent from obeying commands embedded within the ingested data.
Capability inventory: The skill utilizes network access via httpx, database access via asyncpg, and file system read/write access via pathlib to perform its tasks.
Sanitization: The skill demonstrates good practices by using yaml.safe_load() for parsing and by providing a structured validation architecture (validate_records and apply_business_rules) to filter incoming data.

data-ingestion-pipeline