The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local shell script located at .specify/scripts/bash/check-prerequisites.sh during the initialization step to gather project metadata and verify requirements.
[DYNAMIC_EXECUTION]: The skill implements an extension hook system that reads command strings from the .specify/extensions.yml configuration file and dynamically executes them as slash commands (e.g., /speckit-git-commit) within the agent session.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to use external files such as spec.md, plan.md, and tasks.md as the 'sole source of intent'. This creates a surface for indirect prompt injection, as the agent's assessment logic and task generation are directly driven by the content of these external, potentially untrusted documents.
Ingestion points: spec.md, plan.md, tasks.md, and .specify/memory/constitution.md.
Boundary markers: The skill instructs the agent to treat these files as the 'sole source of intent' but does not define specific technical delimiters (like XML tags) for the content being read.
Capability inventory: The skill can execute local bash scripts, trigger other skill commands via hooks, and append data to local files (tasks.md).
Sanitization: The skill includes instructions to escape single quotes when passing arguments to bash scripts.

speckit-converge