palantir-performance-tuning

Pass

Audited by Gen Agent Trust Hub on Sep 12, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill includes functions that ingest data from external Palantir Foundry APIs and datasets, which represents a potential surface for indirect prompt injection if the source data is malicious.\n
  • Ingestion points: Data is fetched via client.ontologies.OntologyObject.list, client.ontologies.OntologyObject.get, and client.ontologies.OntologyObject.search in SKILL.md.\n
  • Boundary markers: No specific delimiters or "ignore instructions" markers are used when processing the data in the provided snippets.\n
  • Capability inventory: The code snippets demonstrate capabilities to read objects, perform batch searches, and write processed results via Palantir Spark transforms.\n
  • Sanitization: The provided patterns focus on performance and do not include explicit sanitization or validation of the content of the retrieved objects.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 12, 2026, 09:36 PM
Security Audit — agent-trust-hub — palantir-performance-tuning