The Agent Skills Directory

[COMMAND_EXECUTION]: Executes shell commands such as git checkout, cat, and git log to manage experiment branches and read configuration files.
[COMMAND_EXECUTION]: Runs a local Python script setup_experiment.py to list experiments and determine their status.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection by reading external data into the agent's context.
Ingestion points: Reads from .autoresearch/{domain}/{name}/config.cfg, program.md, and results.tsv in SKILL.md.
Boundary markers: None identified; files are read directly into the execution context.
Capability inventory: The skill can execute git commands and Python scripts.
Sanitization: No sanitization or validation of the file content is performed before it is presented to the agent.

resume