The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the ao goals CLI suite to perform its core functions. The render --out <file> command specifically allows writing output to a path defined at runtime. Although the instructions contain a warning against pointing this at sensitive source files, this capability represents a surface for arbitrary file writes if the agent's logic is compromised.
[PROMPT_INJECTION]: The skill is subject to indirect prompt injection risks due to its processing of external project files.
Ingestion points: Processes GOALS.md or legacy YAML files from the local filesystem.
Boundary markers: The instructions lack explicit delimiters or warnings to ignore instructions embedded within the processed data.
Capability inventory: The skill can execute CLI commands and perform filesystem write operations.
Sanitization: No sanitization or validation of the input file content is documented before processing or output generation.

fitness