This package is a source-first Markdown archive of all pages listed in Dograh's official documentation index when this archive was created: 124 public documentation pages.1 It is organized to support a later Dograh configuration skill without turning the source material into a lossy summary.
| Use case | Recommended entry point | Why it is useful |
|---|---|---|
| A local agent with a large context window | COMPLETE_DOCUMENTATION.md |
One contiguous export of the official full-documentation feed. |
| A local agent using retrieval or file-based reading | docs/ and CONTENT_MANIFEST.md |
One original Markdown response per official documentation page, plus the raw API schema. |
| A local agent that must inspect the REST schema | OPENAPI_SPECIFICATION.md |
A Markdown wrapper around Dograh's officially indexed OpenAPI JSON schema. |
| A future Dograh skill author | SKILL_FOUNDATION.md |
A focused reading map for setup, configuration, agent building, deployment, and operations. |
| Auditing provenance or updating the archive | ARCHIVE_METADATA.md |
Source URLs, scope, retrieval result, and update guidance. |
The files in docs/ retain the official Markdown returned by each corresponding https://docs.dograh.com/.../*.md endpoint. Original links, code examples, image URLs, and embedded-content references are deliberately retained. This preserves exact documentation content for future skill construction.
The top-level COMPLETE_DOCUMENTATION.md is an official all-pages text export with only a short provenance header added. It is convenient for bulk ingestion but is not a replacement for targeted page retrieval. The official OpenAPI schema is preserved separately both as raw JSON under docs/api-reference/openapi.json and as OPENAPI_SPECIFICATION.md for Markdown-only workflows.
Begin with the entry points in the table below. The detailed, linkable page inventory is in CONTENT_MANIFEST.md.
| Goal | Read these pages first |
|---|---|
| Install or self-host Dograh | docs/getting-started/index.md, docs/getting-started/prerequisites.md, docs/deployment/introduction.md, docs/deployment/docker.md |
| Create an initial voice agent | docs/getting-started/first-agent.md, docs/voice-agent/introduction.md, docs/voice-agent/editing-a-workflow.md |
| Configure models and credentials | docs/configurations/inference-providers.md, docs/configurations/llm.md, docs/configurations/voice.md, docs/configurations/transcriber.md, docs/configurations/api-keys.md |
| Build or run agents programmatically | docs/sdks/introduction.md, docs/sdks/build-an-agent.md, docs/api-reference/agents.md, docs/developer/workflow-schema.md |
| Connect real calling | docs/integrations/telephony/overview.md, followed by the relevant provider page in docs/integrations/telephony/ |
For a future refresh, retrieve the official llms.txt page index and each listed Markdown endpoint again. Compare the resulting URL list with metadata/official_urls.txt to identify additions, removals, or renamed pages. Consult the live documentation before acting on information that may have changed, especially provider capabilities, API behavior, hosted URLs, versions, and deployment instructions.