GoodQA feature guides (agents and integrators)
One page per capability an agent or an integration meets. Each page says whether it is on, who may call it, and links
its API reference, its MCP tools and its runbook. Start with the conversations skill
(served at GET /v1/meta/agent-skill) if you are a bot that talks to people.
| Feature | Guide |
|---|---|
| Agent conversations and delivery | agent-chat.md |
| Ahrefs research and publisher recovery | ahrefs.md |
| Agent Browsers | agent-browsers.md |
| Desktop Agent Browser | desktop-agent-browser.md |
| Calls and meetings | call-intelligence.md |
| Phone call integration | call-intelligence-phone-design.md |
| Browser Control | browser-control.md |
| Cloud workers | cloud-workers.md |
| Personal My work | my-work.md |
| Chat line templates | chat-lines.md |
| Saved CRM reports | crm-reports.md |
Persona event stream (SSE, Last-Event-ID, bus-off 404 + Retry-After) | event-stream.md |
| Persona inbox, poll fallback, webhooks | persona-inbox-and-webhooks.md |
Follow-ups contract (<followups> block, tickets as the record) | followups.md |
| Agent knowledge and persona memory (cursor, match, purge, persona erase, pins) | agent-knowledge.md |
| Shared project memory (provenance, expiry, authority and text/client-vector retrieval) | project-memory.md |
| Vector memory and bring-your-own embedding key | vector-memory.md |
| QA Commander (default persona, desktop, human handoff) | qa-commander.md |
| Custom data (org-owned PostgreSQL) | custom-data.md |
| Import jobs (generic/HighLevel CSV, reviewed chunks, no consent) | import-jobs.md |
| Form to CRM | form-to-crm.md |
| Phone readability | phone-readability.md |
| CRM v2 | crm-v2.md |
| Routines | routines.md |
| Org Blueprints | blueprints.md |
| Coordinator, lanes and handoffs | coordinator.md |
| Agent launches and managed hosting | managed-hosting.md |
| Secret requests | secret-requests.md |
| Credential guard (messages and comments) | credential-guard.md |
| goodqa CLI: terminal chat with QA Commander, vault secrets into one command, custom data | goodqa-cli.md |
| Billing states (grandfathering, seats, not configured, BYOK) | billing.md |
| Usage, cost and agent model spend (idempotent ingest, groupBy model/engine/agent/lane/day, advisory budget) | usage.md |
| Product expert (search every feature, API, tool, command, setting and page, with honest status) | product-expert.md |
| Desktop remote control (Enable remote desktop on this machine, no pairing code, Stop / Block / Turn off) | desktop-remote-control.md |
| One-click Remote (Enable remote for my agent: one consent screen, one record, a visible indicator, local-first Stop, expiry) | one-click-remote.md |
Every MCP tool, with parameters, an example call and related tools: MCP tool reference (generated). Which tools a key gets and when: MCP tools. Who turns what on: feature switches. Docs index.
-
Operations overview — work, named blockers and bot activity across organizations (platform switch off by default).
-
Agent work queue — owner order, pause and liveness.
-
Member setup checklists — admin assignments and server-derived member progress.
-
What’s New and Coming Soon — shipped announcements, private suggestions, unique human votes and explicit agent suggestion grants.
-
Live dashboard work proofs — evidence, expiry and scoped invalidation; parent dashboard gates are unchanged. What is new and coming soon lists release announcements separately from these capability guides.
-
Memory and approved rules: default on; human review and tenant scope apply.
-
Work coordinator — durable ownership, leased workers, review/repair handoffs and daily escalation digests (default on; installation pending).