The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes content from various external project files, creating a potential surface for indirect prompt injection.
Ingestion points: Data is ingested from docs/architecture/tr-registry.yaml, GDD markdown files in design/gdd/, and ADR documents in docs/architecture/ using the Read and Bash tools.
Boundary markers: While the skill organizes output using Markdown sections, it does not implement specific delimiters or safety instructions to prevent the agent from following directions potentially embedded in the ingested documents.
Capability inventory: The skill possesses the ability to modify project files via Write and Edit tools and execute shell commands through the Bash tool (scripts and git).
Sanitization: No explicit sanitization or filtering of the content retrieved from requirement documents is specified before it is incorporated into the agent's context.\n- [DYNAMIC_CONTEXT_INJECTION]: The skill uses the dynamic context injection syntax (!command) in SKILL.md to execute yaml-helper.sh at load time. This operation is used to resolve configuration settings required for the skill's execution from the local environment.\n- [COMMAND_EXECUTION]: The skill executes multiple local shell scripts located in .claude/scripts/ and hooks/ to retrieve story status and metadata. It also uses git commands (git status, git add, git commit) to manage project state and record story completion.

story-done