Transmuting code into confidence — a structured pipeline that orchestrates AI agents across any provider, scores every output, and heals itself when things go sideways.
phases:
spec # define what to build
behavioral # write contracts
adversary # challenge them
tests # prove it works before it exists
implement # write the code
review # score it
test # verify it
transitions:
review.fail: implement # loop back, not forwardTHE PROBLEM
AI agents are coin flips wrapped in confidence. Every run produces different output. Every output needs verification. Most teams either trust blindly or review manually. Neither scales.
The industry built orchestrators for routing — nobody built one for trust.
THE VISION
Orchemist is not another orchestrator. It is a trust engine.
Define what success looks like before writing a single line.
Challenge assumptions before you build on them.
Prove it works before it exists.
Quantify confidence. Gate on thresholds. Route on results.
Retry with feedback, not blindly. Surface blockers, don't hide them.
ARCHITECTURE
A pipeline is a YAML state machine. You define phases, transitions, and quality gates. The engine follows the graph. Swap phases, change models, add gates — without touching code.
SHOWCASES
Spec, adversary, acceptance tests, implement, review, score. Every phase gated. Every output graded.
Built-in anti-hallucination. Every claim gets a source. Red-team review before publish.
Competitive analysis, structured intelligence gathering. From raw data to actionable insight.
Sequential rewrite with flow review and consistency checks. Voice preservation across revisions.
THE MARKET
Seven frameworks compete on how elegantly they route agents. None compete on whether you can trust what those agents produced. LangGraph, CrewAI, Pydantic AI, Google ADK -- all solve communication. Orchemist solves verification.
TRACTION
GET STARTED
Install. Scaffold. Launch. The engine handles the rest.
$ pip install orchemist
$ orch new my-pipeline
$ orch run my-pipeline.yaml --mode openrouter