The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local shell script located at .specify/scripts/bash/check-prerequisites.sh --json to initialize the feature directory and available documentation list. This execution depends on the security of the scripts provided within the project repository.
[DYNAMIC_EXECUTION]: The skill processes .specify/extensions.yml to identify and execute commands defined as extension hooks. Mandatory hooks listed under hooks.before_checklist and hooks.after_checklist are automatically invoked using EXECUTE_COMMAND, allowing for configuration-driven code execution within the agent's environment.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and summarize various requirement files which serves as an attack surface for indirect prompt injection. 1. Ingestion points: Reads spec.md, plan.md, tasks.md, and .specify/extensions.yml from the feature directory and project root. 2. Boundary markers: None present; the skill treats content from requirements files as instructions for generating the checklist without explicit delimiters to prevent command-like behavior in the text. 3. Capability inventory: File reading, directory creation (mkdir), file writing/appending, execution of a local bash script, and execution of arbitrary slash commands defined in a YAML config. 4. Sanitization: Basic shell escaping for single quotes is mentioned for arguments, but there is no evidence of validation or filtering for the content of the ingested requirements documents or hook definitions.

speckit-checklist