The Agent Skills Directory

[SAFE]: The skill serves as a reference and validation layer for NVIDIA TAO artifacts. It defines strict JSON schemas to ensure data integrity and correct structure (e.g., enforcing nested dictionaries and mode discrimination).
[SAFE]: The provided code snippets in SKILL.md and test scripts in references/tests/ use standard Python libraries (jsonschema, pytest, pathlib) for validation purposes. No suspicious network activity, credential handling, or privilege escalation patterns were found.
[SAFE]: The skill.oms.sig file is a standard digital signature bundle (DSSE/Sigstore format) used for verifying the authenticity and integrity of the skill content by NVIDIA. The Base64-encoded certificates within it are part of the public key infrastructure and do not represent malicious obfuscation.
[INDIRECT_PROMPT_INJECTION]: The skill defines a surface for processing untrusted data (TAO artifacts like job records). However, it specifically provides the validation schemas and rules to mitigate malformed or malicious inputs. No exploitable interpolation patterns or missing boundary markers were identified within the skill's logic itself.

tao-artifacts