The Agent Skills Directory

[COMMAND_EXECUTION]: The soc_openroad_run tool executes make commands to automate physical design stages like synthesis, placement, and routing.
[COMMAND_EXECUTION]: The skill supports running OpenROAD flows inside Docker or Podman containers, constructing execution commands that mount the project directory and invoke the flow scripts.
[SAFE]: The server validates design names, module names, and stage targets against expected patterns and whitelists to ensure only valid commands are constructed.
[SAFE]: Command-line arguments for container execution are sanitized using shlex.quote and shlex.join to prevent shell injection vulnerabilities.
[SAFE]: The skill uses subprocess.run with static arguments (--version) to detect the presence of container engines, which is a safe operational check.

soc-openroad