The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it incorporates untrusted data from the repository and user arguments into its planning process.
Ingestion points: The skill processes user-supplied $ARGUMENTS and reads content from arbitrary files in the repository to understand context and map changes.
Boundary markers: The instructions lack explicit boundary markers or directions for the model to ignore potential instructions embedded within the data it reads.
Capability inventory: The skill has the ability to read and write files (creating plans in .agents/plans/ or existing story files) and suggests running terminal commands for verification (lint, typecheck, tests).
Sanitization: There is no evidence of input validation or sanitization of the content retrieved from repository files or user input before it is used to generate the execution plan.

agile-story