Find the right prompt for every model your agent runs on.
Test prompt candidates against objectives and runtime targets, approve the winners, and retrieve the prompt package you should actually use.
Define What "Good" Means
Create objectives with clear success criteria, guardrails, and target runtime matrices. Know exactly what your prompt must achieve.
Evaluate With Evidence
Run deterministic and evaluator-assisted tests against your prompts. Get structured pass/fail evidence, not guesswork.
Resolve With Confidence
Retrieve the Active prompt package for any objective and target runtime. Know which prompt to use, backed by qualification proof.
Developer resources
API and MCP docs for integrating objective workflows into your tools.
Getting started
Set up your first workspace and integration path.
Read guideQuickstart
Run health, auth, and prompt-package calls in minutes.
Start nowMCP and API capabilities
Understand delegated sessions, scopes, and workflow tools.
See capabilitiesOpenAPI reference
Explore endpoint contracts and payload shapes at /api/docs.
Open API docsPublic objective collections
Open objectives the community can inspect, improve, and maintain.