The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by processing external data from the local environment to drive its routing and execution logic.
Ingestion points: The agent reads and interprets the contents of files within the .byte-os/ directory, such as STATUS.md and plans/, to determine the next stage of work.
Boundary markers: The instructions lack explicit delimiters or instructions to ignore embedded commands within the ingested project files.
Capability inventory: The skill possesses capabilities for web browsing (via byte-research fallback) and executing development tasks (via byte-build fallback).
Sanitization: There are no documented procedures for validating or sanitizing the content of the project files before they influence the agent's behavior.

byte-do