The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon data from external sources including validation reports, profile TOML files, and user requests.
Ingestion points: Identification of feature versions and failing requirements occurs via validation reports, profiles, and user-supplied text (as described in the Workflow section of SKILL.md).
Boundary markers: The instructions do not specify the use of delimiters or explicit warnings to ignore embedded instructions within the ingested data.
Capability inventory: The skill involves reading local manifest files (nv_core/...), writing repairs to staged output assets, and executing command-line tools for upgrades and validation.
Sanitization: There is no explicit mention of sanitization or validation of the input data before it influences the agent's actions.
[COMMAND_EXECUTION]: The skill requires the agent to execute shell commands to perform its primary function.
Evidence: The SKILL.md file instructs the agent to use the workspace upgrade command for authoring runtime-specific data and to "rerun the same profile gate or the narrowest available feature/capability validation gate" to verify repairs. While these are intended for local development workflows, they involve executing subprocesses based on configurations derived from potentially untrusted inputs.

simready-foundation-conform-fet-003-physx