The Agent Skills Directory

[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it transforms untrusted user-supplied app ideas into detailed documents and has instructions to save these files to the local workspace.
Ingestion points: The agent takes app concepts or feature lists provided by the user as input for the PRD generation workflow in SKILL.md.
Boundary markers: There are no explicit instructions or delimiters to tell the agent to ignore potentially malicious instructions embedded within the user's input.
Capability inventory: The skill mandates that the agent save the output to the workspace (e.g., docs/PRD.md) as specified in the quality mandates section of SKILL.md.
Sanitization: The skill does not define any validation or escaping mechanisms for user-provided data before it is incorporated into the generated documents.
[EXTERNAL_DOWNLOADS]: The README.md suggests an installation method that fetches the skill from a remote repository (onipinaka/howto-skills) using a command-line tool. This is documented as a vendor-provided distribution mechanism.
[COMMAND_EXECUTION]: The skill instructions in SKILL.md (Mandate 4) direct the agent to perform file system operations, specifically writing markdown files to the project workspace (e.g., PRD_[ProductName]_v1.md), which is the core functionality of the skill.

prd-creator