Plan a coding task, generate and execute Python, and loop on reviewer feedback until the code passes -- a plan/code/judge fix-loop with governed execution.
Research a question with web search and page reads, judge the answer for completeness, and keep digging at the named gaps until nothing major is missing.
Draft an answer, judge it against a rubric, and refine on the judge's written feedback until it passes -- the generate/evaluate/optimize loop.
Draft a plan for a requested action, pause the run for a human approve/reject decision, and only then record the approved plan -- human-in-the-loop approval.
Decompose a task into subtasks with a planner, execute them with a tool-using worker, and synthesize the findings -- the orchestrator/workers pattern, sequential v1.
Answer questions from a document store with citations, then verify groundedness with a judge before accepting -- retrieval-augmented QA that refuses to hallucinate.
Classify an incoming item into urgent/routine/irrelevant and dispatch it to the matching handler via rules -- the LLM emits a label, Fathom routes.
Initialize a new Ruflo project with MCP tools, hooks, and agent configuration