The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process the contents of feature.md and docs/project.md, which creates a surface for indirect prompt injection if those files contain adversarial instructions.
Ingestion points: Instructions in SKILL.md require the agent to read feature.md and docs/project.md in full to identify tech stacks and architecture.
Boundary markers: The instructions do not define clear delimiters or provide the agent with a directive to ignore embedded instructions within the ingested files.
Capability inventory: The skill possesses the capability to write new files (plan.md) to the project root based on the processed data.
Sanitization: There is no evidence of input validation or sanitization of the content read from the Markdown files before it is used to generate the plan.

sdd-plan