AgentRun: DSL to turn agents into workflows

Hi HN,I just open sourced the DSL that our harness in grep.ai uses to turn repeatable parts of agent work into workflows. You can combine tool calls, code, Jev-powered system one decisions for things like routing and screening evidence, and agents when a step needs more investigation.Our harness uses the traces and retro notes agents leave behind when doing a job to figure out which parts can become a workflow. The idea is to make the work easier to understand and avoid paying for a full agent loop where one isn’t needed. For example, a research workflow can split a question into subquestions, send agents to research them in parallel, use Jev to screen the evidence, and have another agent write the report. You can inspect the steps, evaluate the evidence screening separately, or change one agent without rebuilding everything.The DSL and examples are in our GitHub. There’s a scripted demo you can run without API keys: https://github.com/Parcha-ai/agentrunYou can also use it as a Pi extension to build, inspect, and run workflows: https://github.com/Parcha-ai/agentrun#use-it-in-piI would love to hear if this is useful to others.More background on how AgentRun works in this video: https://www.youtube.com/watch?v=vOVhtGjtwpg. Or read about our use cases in this article: https://x.com/MiguelriosEN/status/2101029313906987422.

Hacker NewsToolAISource
0Sign in to voteCopy link

FL score

62

out of 100

Verdict

VALIDATE

high confidence

Competition

No competitor data yet

Trend

No signal yet

A DSL that converts repetitive agent work into cheaper, inspectable workflows by analyzing agent traces.

The pain

Teams using AI agents pay for full LLM loops on tasks that could run as deterministic workflows. Agent outputs are hard to debug, modify, or optimize. Costs scale poorly when agents repeat similar reasoning patterns.

The gap

Existing agent frameworks like LangChain and AutoGen focus on building agents, not optimizing them. No standard tool exists to extract workflow patterns from agent behavior and convert them to cheaper execution paths. LLM providers have not solved this because it reduces token consumption.

Build angle

Start by targeting research and data extraction teams who run agents repeatedly. Charge per workflow conversion or per execution savings. Build integrations with popular agent frameworks. Create a marketplace for pre-built workflows to increase stickiness.

Strengths

  • Solves a real cost problem as agent usage scales.
  • Working prototype with open source code and demo.
  • Addresses a gap between agent builders and agent optimizers.
  • Inspection and debugging features add value beyond cost savings.
  • Parallel execution and routing logic are genuinely useful patterns.

Risks

  • Market may be too early. Most teams still experimenting with agents, not optimizing them.
  • LLM providers could build this natively into their platforms or agent frameworks.
  • Unclear how to monetize without becoming a cost center that customers resent.
  • Requires deep integration with customer workflows, making sales cycles long.
  • Open source strategy may cannibalize commercial opportunity if adoption stays hobbyist.
  • Competitive threat from larger orchestration platforms adding workflow features.
  • Value proposition depends on agent costs staying high, which may not hold as models improve.

Questions about this idea?

FlyBot reads the scoring and gives you a second opinion on “AgentRun: DSL to turn agents into workflows”.

Open FlyBot