The Agent Skills Directory

[COMMAND_EXECUTION]: The skill automatically executes agent commands defined in the .specify/extensions.yml configuration file. It identifies mandatory hooks (where optional is false) and uses the EXECUTE_COMMAND instruction to run them automatically during the implementation process.
[COMMAND_EXECUTION]: The skill runs a PowerShell script (.specify/scripts/powershell/check-prerequisites.ps1) located within the repository root. Executing local scripts from an external repository is a significant risk as the code within the script is not verified.
[DYNAMIC_EXECUTION]: Command identifiers are dynamically constructed at runtime by reading strings from the project's YAML configuration and transforming them (e.g., replacing dots with hyphens) before execution.
[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it parses and acts upon instructions and metadata found in multiple repository files without sanitization.
Ingestion points: .specify/extensions.yml, tasks.md, plan.md, and project documentation files.
Boundary markers: None identified; the skill treats the content of these files as authoritative instructions for its implementation logic.
Capability inventory: File system modification (creating/updating ignore files), execution of local PowerShell scripts, and automated invocation of other agent commands.
Sanitization: The skill does not perform validation or sanitization on the command names or task descriptions extracted from the project files before acting on them.

speckit-implement