The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the execution of shell commands through the run_shell_command function in scripts/common.py. This capability is used by scripts/run_experiment.py to perform benchmark trials and validation checks defined in the experiment configuration.
[PROMPT_INJECTION]: The SKILL.md file contains instructions that direct the agent to adhere to a structured 'experiment loop,' including mandatory up-front Q&A and hypothesis documentation, which serves as a behavioral constraint for the agent's operation.

autoresearch