The Agent Skills Directory

[SAFE]: The skill is restricted to text processing and documentation management. It reads project-specific files to inform its output and writes the resulting roadmap to the local file system.
[SAFE]: All file operations are localized to the "docs/product/" directory. No sensitive system files, environment variables, or credentials are accessed or exposed.
[SAFE]: No network-based exfiltration or external downloads are attempted by the instructions or metadata. All data flows are internal to the workspace.
[SAFE]: The skill does not utilize any dynamic code execution, command-line subprocesses, or obfuscation techniques.
[SAFE]: Regarding indirect prompt injection attack surfaces:
Ingestion points: The skill reads "docs/product/.ompm/prioritization.json" and "docs/product/positioning.md" (SKILL.md).
Boundary markers: None identified in the skill instructions.
Capability inventory: The skill is capable of writing files to "docs/product/roadmap.md" and "docs/product/.ompm/roadmap.json" (SKILL.md).
Sanitization: No explicit sanitization or validation of input data is mentioned. The risk is minimal as the skill lacks high-privilege capabilities such as shell execution or network access.

roadmap-planning