Set task success flag based on actual task result status (#39)

Fix browser config to browser profile (#37)

Instruct planner to launch first of n tasks upon start (#36)

Update tool to use the first successful result (#35)

Improve task parallelism planning and execution (#34)

Support launching executors in parallel (#33)

Support enabling evaluation for individual tasks (#32)

Merge pull request #31 from stanford-mast/rw-update-eval-readme

Improve instructions for setting up custom eval

Support setting allowed domains for eval

Support setting allowed domains for eval

Merge pull request #30 from stanford-mast/rw-support-custom-task-eval

Improve logger message

Save final state path

Save final state path