# canonic

> The Open Context Layer for Data Agents

## Docs

- [canonic completion](https://docs.getcanonic.app/cli-reference/completion.md): Generate a shell completion script.
- [canonic connection](https://docs.getcanonic.app/cli-reference/connection.md): Manage source connections in canonic.yaml.
- [canonic eval](https://docs.getcanonic.app/cli-reference/eval.md): Evaluate local models against the tested baseline.
- [canonic ingest](https://docs.getcanonic.app/cli-reference/ingest.md): Reconcile configured sources into reviewable context diffs.
- [canonic knowledge](https://docs.getcanonic.app/cli-reference/knowledge.md): Search and author knowledge pages.
- [canonic mcp](https://docs.getcanonic.app/cli-reference/mcp.md): Control the local MCP daemon.
- [CLI Overview](https://docs.getcanonic.app/cli-reference/overview.md): Global flags and how the canonic CLI is organized.
- [canonic query / sql / assert](https://docs.getcanonic.app/cli-reference/query-sql-assert.md): Execute queries, run raw read-only SQL, and gate on accuracy.
- [canonic review / apply](https://docs.getcanonic.app/cli-reference/review-apply.md): Review and apply pending proposals from an ingest run.
- [canonic setup](https://docs.getcanonic.app/cli-reference/setup.md): Interactive project setup wizard.
- [canonic sl](https://docs.getcanonic.app/cli-reference/sl.md): Semantic layer: resolve, compile, and describe.
- [canonic status / report](https://docs.getcanonic.app/cli-reference/status-report.md): Inspect project state and local event-log figures.
- [The semantic compiler](https://docs.getcanonic.app/concepts/compiler.md): Deterministic, LLM-free translation from a semantic query to dialect-correct SQL.
- [Connectors](https://docs.getcanonic.app/concepts/connectors.md): How canonic reads a warehouse, a dbt project, a BI tool, or a doc: through one capability contract.
- [Contracts & guardrails](https://docs.getcanonic.app/concepts/contracts-and-guardrails.md): What's canonical, what's mandatory, and what the compiler must obey.
- [Ingestion & reconciliation](https://docs.getcanonic.app/concepts/ingestion-and-reconciliation.md): How normalized evidence turns into reviewable diffs, never a silent edit.
- [Instrumentation & evaluation](https://docs.getcanonic.app/concepts/instrumentation-and-eval.md): The local event log, the accuracy harness, and the model baseline: how canonic measures itself.
- [Knowledge](https://docs.getcanonic.app/concepts/knowledge-layer.md): Free-form, searchable, auto-maintained pages that carry business meaning.
- [LLM & embeddings runtime](https://docs.getcanonic.app/concepts/llm-runtime.md): One interface to generation and embeddings; the enforced air-gapped guarantee.
- [Semantics](https://docs.getcanonic.app/concepts/semantics.md): How canonic describes a table well enough to query it safely.
- [The three layers](https://docs.getcanonic.app/concepts/three-layers.md): Semantics, knowledge, and contracts, and the split rule for where new context belongs.
- [Configuring an LLM](https://docs.getcanonic.app/configuring-an-llm.md): canonic's answer path is fully deterministic; an LLM is optional and only used to draft context.
- [Guide: Dutch Railway Network (DuckDB)](https://docs.getcanonic.app/guides/dutch-railway.md): A DuckDB-native example with a geography dimension chain and ratio metrics.
- [Guide: Ecommerce (Postgres)](https://docs.getcanonic.app/guides/ecommerce.md): A small but complete Postgres project covering the full Phase 1 loop, evidence connectors, and observability.
- [Guide: Jaffle Shop (dbt + MetricFlow)](https://docs.getcanonic.app/guides/jaffle-shop.md): The classic dbt Jaffle Shop dataset, with MetricFlow semantic models as modeling-tier evidence.
- [Guide: Vehicle Rental (SQLite)](https://docs.getcanonic.app/guides/rental.md): A SQLite-only project with cross-fact fanout and a NULL-handling guardrail.
- [Guide: SaaS Analytics (all 7 binding kinds)](https://docs.getcanonic.app/guides/saas-analytics.md): The broadest example: every metric binding kind, all three guardrail kinds, finality, and assertions.
- [Installation](https://docs.getcanonic.app/installation.md): Install the canonic CLI via npm, Homebrew, or Docker.
- [Introduction](https://docs.getcanonic.app/introduction.md): The context layer that lets AI agents query your data correctly.
- [Connecting your agent](https://docs.getcanonic.app/mcp-integration/connecting-your-agent.md): Expose canonic's capabilities to an agent client over MCP.
- [MCP tools reference](https://docs.getcanonic.app/mcp-integration/tools-reference.md): The tools canonic's MCP server registers, grouped by purpose.
- [Quickstart](https://docs.getcanonic.app/quickstart.md): Your first answer in minutes.
- [canonic.yaml config schema](https://docs.getcanonic.app/reference/config-schema.md): Every top-level field in the project config file, and what it governs.
- [Error codes](https://docs.getcanonic.app/reference/error-codes.md): The canonical error registry and its headless exit-code mapping.