The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is instructed to run a local shell script located at .specify/scripts/bash/check-prerequisites.sh from the repository root.
[DYNAMIC_EXECUTION]: The skill implements a 'hooks' mechanism that reads command names from a local YAML file (.specify/extensions.yml), converts them into slash commands, and instructs the agent to 'actually invoke the hook and wait for it to finish'. This allows the repository to define arbitrary commands that the agent will execute, creating a significant attack surface for local command execution.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data that could contain malicious instructions.
Ingestion points: .specify/extensions.yml (commands and prompts) and tasks.md (task descriptions).
Boundary markers: No explicit markers or 'ignore' instructions are provided when processing the task list or executing hooks.
Capability inventory: The agent has the ability to run shell scripts, execute project-defined commands, and interact with the GitHub API to create and list issues.
Sanitization: No sanitization is mentioned for the content extracted from tasks.md before it is used to generate GitHub issue titles and descriptions.

speckit-taskstoissues