The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-controlled artifacts such as spec.md and plan.md to coordinate development tasks, creating a surface for potential indirect prompt injection attacks.
Ingestion points: The skill reads project-specific artifacts including .specify/memory/constitution.md, spec.md, plan.md, and tasks.md during its initialization and execution phases.
Boundary markers: The workflow relies on structured markdown templates and defined artifact sections (as documented in modules/artifact-structure.md) to organize and delimit content.
Capability inventory: The skill's capabilities are focused on file system operations within the local project repository (reading and writing artifacts in .specify/ and specs/ directories). It does not perform network operations or execute arbitrary shell commands outside this scope.
Sanitization: The orchestrated workflow includes explicit verification checkpoints and quality checklists (e.g., modules/progress-tracking.md defines completion verification patterns) to validate the content and structure of artifacts before proceeding to implementation.

speckit-orchestrator