DOCS / LEARN

Cookbooks

Learn the framework through reproducible, synthetic incident investigations with explicit safety boundaries.

PHASE 1 · PRE-1.0PYTHON 3.11+EDIT ON GITHUB ↗

Available cookbooks

Every cookbook name links to its directory on GitHub, where the README walks through setup and each step. All of them include an offline smoke path—no credentials or network required.

CookbookWhat it demonstrates
Simple log diagnosisComplete offline CLI flow from a local failure log to report and memory.
Structured rule evaluationTyped fields, quantifiers, required evidence, ranking, and fixture testing.
Evidence and JSON reportingBounded evidence collection and the versioned machine-readable report.
Prometheus / AlertmanagerDiagnosing a real-world-shaped alert webhook with rules and bounded evidence.
Data pipeline investigationSchema, lineage, telemetry, code, and knowledge context.
ML regression monitoringFeature drift and model-performance investigation.
Software-delivery CI investigationDependency, permission, and infrastructure-reference investigation.
Recording a resolutionHuman-confirmed resolution and the local truth transition.
Guarded proposalPlaybooks, default-deny policy, evidence-linked proposals, and approvals.
Verification and replayExplicit verification records, conservative learning, and replay evaluation.
Webhook and HTTP evidenceAuthenticated webhook normalization and the hardened HTTP evidence connector.
PostgreSQL memoryShared durable memory and custom schemas through the independent plugin.
Plugin packageA complete independently packaged plugin with manifest and contract tests.

Framework versus cookbook

The core owns typed interfaces and safe defaults. Each cookbook owns its synthetic fixtures, scenario rules, knowledge, service composition, prompts, and any optional Agno or OpenRouter integration. Start with a cookbook for a runnable demonstration, then use the architecture and reference pages to examine the framework contracts behind it.