# Synkro

## Docs

- [Advanced Module](https://synkro.sh/docs/advanced/index.md): Direct access to internal components for power users
- [create_pipeline()](https://synkro.sh/docs/api-reference/create-pipeline.md): Create a customized generation pipeline
- [Dataset](https://synkro.sh/docs/api-reference/dataset.md): Dataset class for managing generated traces
- [Formatters](https://synkro.sh/docs/api-reference/formatters.md): Export formats for training datasets
- [generate()](https://synkro.sh/docs/api-reference/generate.md): Generate training datasets from policy
- [generate_scenarios()](https://synkro.sh/docs/api-reference/generate-scenarios.md): Generate eval scenarios without synthetic responses
- [grade()](https://synkro.sh/docs/api-reference/grade.md): Grade a response against policy compliance
- [Model Constants](https://synkro.sh/docs/api-reference/models.md): Supported LLM providers and model constants
- [Overview](https://synkro.sh/docs/api-reference/overview.md): Synkro API reference
- [Policy](https://synkro.sh/docs/api-reference/policy.md): Policy class for managing policy documents
- [Reporters](https://synkro.sh/docs/api-reference/reporters.md): Progress reporting classes for generation pipeline
- [Session](https://synkro.sh/docs/api-reference/session.md): Database-backed sessions with HITL methods
- [simulate()](https://synkro.sh/docs/api-reference/simulate.md): Run auto-generated scenarios against a live agent
- [Types](https://synkro.sh/docs/api-reference/types.md): All public type definitions in synkro
- [Coverage Tracking](https://synkro.sh/docs/concepts/coverage.md): Ensure your dataset covers all policy aspects
- [Logic Map](https://synkro.sh/docs/concepts/logic-map.md): Structured rule extraction from your policies
- [Pipeline](https://synkro.sh/docs/concepts/pipeline.md): How Synkro transforms documents into datasets
- [Conversation](https://synkro.sh/docs/datasets/conversation.md): Multi-turn dialogue datasets for chat models
- [Evaluation](https://synkro.sh/docs/datasets/evaluation.md): Q&A datasets with ground truth for LLM evaluation
- [Instruction](https://synkro.sh/docs/datasets/instruction.md): Single-turn Q&A datasets for instruction-following models
- [Tool Calling](https://synkro.sh/docs/datasets/tool-call.md): Train models to use custom tools
- [Examples](https://synkro.sh/docs/examples/index.md): Copy-paste ready code examples for common workflows
- [Agent Simulation](https://synkro.sh/docs/guides/agent-simulation.md): Test your agent against auto-generated scenarios grounded in policy rules
- [Checkpointing](https://synkro.sh/docs/guides/checkpointing.md): Resume interrupted generation jobs
- [Cost Tracking](https://synkro.sh/docs/guides/cost-tracking.md): Monitor LLM costs and API calls during generation
- [Coverage Improvement](https://synkro.sh/docs/guides/coverage-improvement.md): Optimize scenario coverage across policy sub-categories
- [Human-in-the-Loop Workflow](https://synkro.sh/docs/guides/hitl-workflow.md): Interactive editing of Logic Maps and scenarios
- [HuggingFace Integration](https://synkro.sh/docs/guides/huggingface-integration.md): Push datasets to HuggingFace Hub
- [Tool Calling Datasets](https://synkro.sh/docs/guides/tool-calling.md): Generate training data for function calling models
- [Installation](https://synkro.sh/docs/installation.md): Install Synkro and configure your environment
- [Introduction](https://synkro.sh/docs/introduction.md): AI agent simulation framework — test and ship with confidence
- [Quickstart](https://synkro.sh/docs/quickstart.md): Generate your first dataset in 5 minutes
- [Quick Reference](https://synkro.sh/docs/reference/cheatsheet.md): Copy-paste patterns for common synkro operations
- [Parameter Reference](https://synkro.sh/docs/reference/parameters.md): Complete parameter reference for all functions

## OpenAPI Specs

- [openapi](https://synkro.sh/docs/api-reference/openapi.json)