The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute local bash scripts, specifically scripts/set-feature-record-lifecycle.sh, as part of its workflow.
[COMMAND_EXECUTION]: The skill directs the agent to run a "checker command" that is dynamically defined within the config/change-contract-policy.sh file. This pattern allows for the execution of arbitrary commands based on the contents of a local configuration file.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes and acts upon untrusted data from the repository, such as policy files and feature records.
Ingestion points: config/change-contract-policy.sh and files within the feature_records/ directory (identified in SKILL.md).
Boundary markers: No delimiters or instructions to ignore embedded commands within the processed files are specified.
Capability inventory: Shell command execution via bash and dynamic execution of commands sourced from configuration files.
Sanitization: There is no evidence of sanitization, validation, or escaping of the content read from the repository files before it influences agent actions.

frame-development-contract