The Agent Skills Directory

[COMMAND_EXECUTION]: The activation protocol includes the execution of a local Node.js script: node .aiox-core/development/scripts/generate-greeting.js dev. This occurs automatically when the skill is initialized.
[COMMAND_EXECUTION]: The skill provides the *verify-subtask command, which allows the agent to execute arbitrary system commands, API requests, and browser-based end-to-end tests to verify implementation tasks.
[PROMPT_INJECTION]: The skill utilizes persona-adoption instructions ('Adopt this agent persona and command system'). While standard for defining agent behavior, this is a form of prompt manipulation.
[PROMPT_INJECTION]: The skill processes potentially untrusted data from local project artifacts to drive autonomous behavior.
Ingestion points: The agent reads instructions and task definitions from .aiox-core/development/agents/dev.md, implementation.yaml, and recovery/attempts.json.
Boundary markers: No specific delimiters or safety instructions are defined to separate data read from these files from the agent's internal control logic.
Capability inventory: The agent possesses extensive capabilities including Node.js execution, shell command execution, and autonomous file modification (*develop-yolo).
Sanitization: The instructions do not specify any validation or sanitization of the content extracted from configuration files before it is processed by the agent.

aiox-dev