The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/check-readiness.py executes the git command using subprocess.check_output to obtain the repository's HEAD revision. The path passed to the git -C command is derived from the location of the input readiness JSON file.
[DYNAMIC_EXECUTION]: In scripts/check-readiness.py, the skill attempts to execute an external Python script (check-profile.py) located at a relative path (../../profile-system-architecture/scripts/check-profile.py) outside of its own directory. This behavior depends on the filesystem layout and executes code from a computed path.
[INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection as it processes untrusted JSON data to make automated 'GO/NO-GO' decisions for system deployments.
Ingestion points: The scripts/check-readiness.py script reads and parses user-supplied JSON files (readiness and architecture profiles).
Boundary markers: The script checks for independent reviewers but does not use explicit prompt delimiters when processing the content of the evidence files.
Capability inventory: The script has capabilities to execute shell commands (git) and run other Python scripts.
Sanitization: The script performs SHA256 checksum verification on artifacts and strictly validates the JSON structure against expected schemas, which significantly limits the risk of traditional injection attacks.

assess-replacement-readiness