The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructs the agent to perform autonomous system changes when specific errors occur. Specifically, it states: "invoke the revx-auth skill immediately and apply the fix yourself. Do not interrupt the user to ask them to run setup." This encourages the agent to bypass human review for configuration and security-related fixes.
[INDIRECT_PROMPT_INJECTION]: The skill establishes a monitoring loop where the agent reads the output of a background process. This creates a vulnerability surface where malicious content embedded in external data (such as exchange messages, symbols, or error logs) could influence the agent's behavior.
Ingestion points: Reading background task output files generated by the revx strategy grid run command (SKILL.md).
Boundary markers: None mentioned; the agent is simply told to "report key events to the user".
Capability inventory: Uses Bash for execution, Read for file monitoring, and TaskStop for process management.
Sanitization: No instructions provided to sanitize or validate the content of the monitored logs before processing.
[COMMAND_EXECUTION]: The skill heavily relies on the Bash tool to execute various subcommands of the revx CLI (backtest, optimize, run). While this is the primary purpose of the skill, it represents a broad execution capability within the user's environment.

revx-strategy