Skip to content

Yareta User & Operational Documentation

How people use, operate, support, and maintain the Yareta platform. This site covers the User & Operational Documentation category only.

Technical & architecture documentation (system architecture, API/DB docs, Mermaid architecture/ER/UML/data-flow diagrams, implementation notes, chat-sparkle-vision docs, known code-vs-doc gaps) lives in the sibling repo yareta-technical-documentation, not here.

It was generated from a source-code audit in August 2026 — every claim is grounded in the actual codebase, not generic boilerplate.

Documents

DocumentCovers
QuickstartSign up → add your first deal → see your first score, in six steps
User GuideHow end users use the application
GlossaryEvery product term (DNA Score, Traffic Light, Competency DNA, and so on) in one place
Integrations Setup GuideConnector-by-connector setup steps (Airtable, HubSpot, Fireflies, and the rest)
Admin GuideAdmin console functionality and configuration
Operations RunbookWhat to do when something goes wrong
Deployment / Release RunbookStep-by-step release and rollback process
Incident Response GuideHandling production incidents
Troubleshooting GuideCommon issues and how to resolve them
Support / FAQCommon user/support questions
Onboarding GuideHow a new developer/operator gets started
Maintenance GuideRoutine maintenance, upgrades, cleanup, rotations
Business Continuity PlanHow the business/service continues during major failures
Security & Privacy OverviewCustomer-facing security/privacy draft — needs legal review before external use

The Five Things Worth Knowing Before You Read Anything Else

  1. The GraphQL endpoint is POST /api, not /api/graphql — a common mistake, including in this repo’s own spectaql.yml.
  2. schema.graphql is stale. Several live modules aren’t in the checked-in file. Regenerate via scripts/generate-schema.ts before trusting generated API docs.
  3. Editing a prompt in code doesn’t change production behavior until it’s synced to the system_prompts database table. This has caused at least one real production incident.
  4. Two deployable units, one shared database. The Next.js app (Cloudflare Pages) and the background job worker (a separate Cloudflare Worker) deploy independently but read/write the same D1 database per environment.
  5. Migration filenames collide. 24 numeric migration prefixes have more than one .sql file on disk. Check the current highest number before generating a new migration, and never renumber the existing ones.

For the fuller, code-cited list of known drift and security-relevant gaps (including several not covered above — GraphQL auth surface, OTP bypass mechanism, Lovable scoring-constraint violations), see yareta-technical-documentation/technical-docs/KNOWN-GAPS.md in the sibling repo.

Known Documentation Gaps

Two things this audit identified but deliberately did not fabricate content for, rather than publishing something misleading:

  • No changelog / release notes. There’s no version history or “what’s new” page for end users, and no changelog data in the codebase to source one from. If this is wanted, it needs a process going forward (e.g. a short entry per customer-visible release) rather than a backfilled one.
  • The Security & Privacy Overview above is a first draft, not an approved policy — it’s flagged as such at the top of that document. It should not be sent to a customer or used in a security questionnaire until someone who owns that function has reviewed it.

Maintaining This Documentation

Treat staleness here as a demonstrated, recurring risk, not a hypothetical one. When you make an operationally significant change, update the relevant file in the same PR. See the Maintenance Guide’s documentation-currency section.