The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by ingesting and processing untrusted data from SQL files and database schemas.
Ingestion points: The workflow uses glob and read to ingest SQL content from models/**/*.sql and uses schema_inspect and schema_detect_pii to process database schema metadata.
Boundary markers: The instructions do not specify boundary markers or explicit instructions for the agent to ignore potentially malicious content embedded within the analyzed SQL files.
Capability inventory: The skill utilizes read-only discovery and analysis tools (glob, read, altimate_core_classify_pii, altimate_core_query_pii, schema_inspect). No capabilities for file writing, network exfiltration, or arbitrary command execution were found.
Sanitization: There is no documented evidence of sanitization or filtering applied to the SQL content before it is processed by the analysis tools.

pii-audit