How it's built · the engineering

You've seen it work.
Here's how it's built.

A workflow-automation engine written from scratch — no orchestration framework underneath. Single-purpose specialists with fixed contracts, composed into validated control-flow graphs, run by a concurrency-safe dispatcher, with human gates and a learning loop. Every decision below is one I can defend end to end.

specialists + contracts control-flow graphs static validator the dispatcher learning loop per-tenant isolation

The reusable parts — one job, reused everywhere

One specialist, three jobs

verify runs in all three recipes. Its job never changes — does this output actually hold up? — only the standard it's checked against differs. Same code, three configs; that's why a new use case mostly reuses what's on the shelf instead of building from scratch.

verify · in triage

“the classification and routing fit the issue, and the reply is relevant and asks for anything genuinely missing”

pass · 0.60 — flagged: the reply never asks for the Turbopack / Next.js version, which actually matters here.

verify · in content

“on-brand (no hype, no emojis, numbers over adjectives), coherent, and not making unsupported factual claims”

pass · 0.82 — flagged: “numbers over adjectives” barely exercised — the piece stays conceptual.

verify · in refine

STRICT — ≥ 2 concrete numeric stats · names a specific failure mode · no hype words · no emojis · under 4 sentences

pass · 0.90 — no issues; this is the bar the draft→rewrite loop chased until it cleared.

Same specialist, three jobs — the recipe author sets the standard, the specialist does the rest. Most parts across these recipes are shared exactly like this; only the recipe-specific middle is new.

Under the hood — three parts, click to open

How it works

Describe a task → the Composer builds a validated workflow → the Dispatcher runs it on a board → the Specialists do the work and learn. One line each; open any for the detail.

Given only a plain-English description, the Composer re-wrote specialists I'd hand-built — scored against their real, human-labeled evals, it matched the originals. When it can't cover a use case it flags the missing specialist and refuses to call the recipe runnable — it doesn't fake capability. The Validator then statically checks the graph (entry/exit, reachability, loop termination, contract compatibility) and rejects invalid ones up front, not halfway through a run.

specialistauto-writtenhand-built
classify6/6 · 100%6/6
rank3/3 · 100%3/3
route3/3 · 100%3/3

The board is the shared surface the dispatcher writes onto. Work streams in and flows through the graph concurrently; a specialist clones per item (the stacked chips); when verify flags a draft it loops back to rewrite (the red curve); the rest wait at the gate for a human. Deliberately dumb, single-writer, concurrency-safe by construction — the piece I'm most proud of. Flow is illustrative.

content · board
items flow in continuously · specialists clone under load · all pause at the gate
in flight flagged → loops back to rewrite held at the gate approved → output

Each specialist is a tiny agent with one fixed contract, one job — small enough to test and trust on its own. And it learns: a deliberately weak one went 4/8 → 8/8 from six curated examples on an independent eval (retrieval few-shot, no fine-tuning), then was promoted to trusted. the real run →

it learns — measured
cold
+6 examples

Cold, it follows a wrong instinct. Six curated examples → 50% → 100%, then trusted. Retrieval-injected few-shot, no fine-tuning.

the contract — one job, config-driven
classify  # item → one category
  in   { item, categories, criteria }
  out  { label, confidence, reasoning }

The same instance labels bug types for triage and content angles for content — only the categories change. That's why parts are reusable.