The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from requirements.md, design.md, and various project documentation files to generate implementation tasks, creating a potential vector for indirect prompt injection where malicious instructions in these documents could influence the implementation plan.
Ingestion points: The skill reads data from requirements.md, design.md, docs/agents/project.md, docs/standards/, and docs/architecture/ files.
Boundary markers: The instructions do not define the use of specific delimiters or protective prompts to prevent the agent from following instructions embedded within the processed requirement or design files.
Capability inventory: The skill has the ability to write implementation plans to the file system, call other execution skills (build-in-waves, build-by-story, build-inline), and interact with external issue trackers.
Sanitization: No explicit sanitization, filtering, or validation of the ingested requirement or design content is mandated before it is used to define the tasks and commands in the implementation plan.

plan-tasks