The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local PowerShell script (.specify/scripts/powershell/check-prerequisites.ps1) found within the user's repository. This pattern allows for the execution of arbitrary code if a malicious script is placed at that specific path in a processed project.
[DYNAMIC_EXECUTION]: The skill implements a 'hooks' system that parses .specify/extensions.yml and triggers commands via EXECUTE_COMMAND: {command}. This allows external configuration files within the repository to dynamically control agent behavior and execute potentially sensitive slash commands.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes multiple files from the local repository that could contain malicious instructions.
Ingestion points: .specify/extensions.yml, .specify/memory/constitution.md, the feature specification file, and FEATURE_DIR/checklists/requirements.md.
Boundary markers: None present. The instructions do not define delimiters or warnings to ignore embedded instructions within these ingested files.
Capability inventory: The skill has the ability to execute PowerShell scripts, trigger dynamic hook commands, and overwrite local files (specs and checklists).
Sanitization: No sanitization or validation of the content within the ingested markdown or YAML files is performed before processing.

speckit-clarify