The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a PowerShell script located at .specify/scripts/powershell/check-prerequisites.ps1 from the repository root. Executing local scripts based on project structure involves risk if the project files are from an untrusted source.
[DYNAMIC_EXECUTION]: The skill implements an extension hook mechanism that reads command names and parameters from .specify/extensions.yml and executes them. It dynamically constructs command identifiers (e.g., transforming speckit.git.commit to /speckit-git-commit) and instructs the agent to invoke these commands automatically, which constitutes execution of instructions from a mutable local configuration file.
[INDIRECT_PROMPT_INJECTION]: The skill is vulnerable to instructions embedded in processed data files.
Ingestion points: Reads content from tasks.md, .specify/extensions.yml, and .specify/memory/constitution.md.
Boundary markers: None identified; the skill does not use specific delimiters or instructions to ignore embedded commands in the ingested data.
Capability inventory: Can execute shell commands (git), run PowerShell scripts, and create GitHub issues via the GitHub MCP server.
Sanitization: No sanitization or validation of the ingested task descriptions or hook prompts is performed before they are processed or used in further commands.

speckit-taskstoissues