Back to directory

DeterminFlow

A production-oriented AI workflow runtime that organizes language-model calls, scripts, APIs, database actions and human approvals into versioned, retryable and auditable workflows, instead of one long-running single agent.

DeterminFlow

DeterminFlow is a workflow runtime for running complex AI processes as dependable services rather than as one long single-agent conversation. Each agent in a workflow handles one clearly bounded node: it reads only the context that step needs, uses only the tools it is authorized to use, and delivers a result the runtime can validate, while DeterminFlow itself manages control flow, data flow, retries, recovery and auditing across the whole workflow.

Compared with handing an entire process to a single-agent framework such as Codex or Claude, changing a workflow here means editing versioned nodes, variables, branches and subprocesses rather than prompt text; each agent node keeps its own isolated context instead of an ever-growing history; a failed node can be retried or resumed without rerunning completed work; and tool permissions, structured output validation and per-node cost accounting are scoped to individual nodes. The authors report a real production case, an AI novel-writing pipeline, that used a large majority fewer tokens than an equivalent single-agent approach.

It includes a visual workflow editor, FastAPI and React components, scheduled automation, and packaging as an API, background service or plugin.

More in Agent frameworks

See all 902
  • Recurse

    recurse.run

    A serverless harness that lets a coding agent design, test, and deploy its own specialist sub-agents as tools, MCP servers, or bots, defined by a schema-validated manifest.

  • ProofForge

    shulgin.is-a.dev

    An AI-agent pipeline that decomposes a mathematics problem, proves each piece, and formalizes the result in Lean 4 and Mathlib until the Lean kernel accepts it, with six of its proofs already merged into Google DeepMind's formal-conjectures repository.

  • local-coder

    github.com

    Builds a local coding-agent team on top of Ollama and OpenCode by detecting the machine's hardware, recommending models for seven specialist roles, and testing whether each assigned model can actually use tools, edit a repository and pass verification.