Skip to main content
Woes brings your support conversations, API knowledge, and AI support workflow into one workspace. Use it to answer developer questions from live chat and Discord, ground the AI agent in your API context, test requests with workspace auth, and hand sensitive or uncertain cases back to your team. The product is organized around one conversation model. Live chat, Discord, and future email intake are channels on the same inbox rather than separate tools. The AI agent is a participant in that support workflow: it can draft or respond when there is enough grounded context, and it should clarify or hand off when confidence is low.

​
What you can do with Woes

  • Manage customer conversations from live chat and Discord in a shared operator inbox.
  • Add API documentation, OpenAPI specs, Markdown, JSON, and website sources to API Context.
  • Review generated endpoint summaries and source chunks before relying on AI answers.
  • Configure API credentials separately from source content, then run auth-aware API tests.
  • Use the Lab to test agent answers, citations, confidence, retrieval, and live API verification.
  • Install the customer widget on your site with a workspace public key.
  • Track issues, notes, tags, custom fields, macros, and automations around support work.
Woes is designed for grounded developer support. The agent should not invent endpoints, fields, auth rules, responses, or production behavior that are not present in your workspace context.

​
Main areas

Public API

Preview the planned public API categories for Woes integrations.

Quickstart

Set up your workspace, add API context, test the agent, and open your first support channel.

Workspace setup

Configure members, roles, tags, custom fields, macros, automations, and workspace settings.

API context

Ingest API sources, review generated endpoint knowledge, attach credentials, and run tests.

Agent and Lab

Understand how the AI agent retrieves evidence, cites sources, handles confidence, and hands off.

Channels and inbox

Work live chat and Discord conversations from one shared operator queue.

Live chat widget

Install the widget, identify users, pass display context, and handle conversation history safely.

Security and privacy

Learn the boundaries around tenant isolation, credentials, public keys, redaction, and rate limits.
1

Create your workspace

Invite your team, assign roles, and add the fields, tags, and macros operators need before conversations start arriving.
2

Add API sources

Upload or connect your API docs in API Context. Scan each source, then inspect the generated endpoint and Markdown evidence.
3

Configure safe API testing

Add credentials separately from documentation content. Test a small set of endpoints before enabling agent workflows that depend on live verification.
4

Test the agent in Lab

Ask real support questions, check citations, verify confidence behavior, and confirm unclear questions are clarified or handed off.
5

Open channels

Install the live chat widget or connect Discord, then handle customer messages from the shared inbox.

​
Current product boundaries

  • Live chat and Discord are active inbound support channels.
  • Email configuration currently supports verified domains and outbound testing. Treat inbound email intake as a roadmap area unless your workspace has been explicitly enabled for it.
  • The browser widget public key is intentionally public. It identifies the workspace but does not grant general data access.
  • Conversation history in the widget requires a per-conversation history secret. Verified user identity is separate and does not replace that secret.
  • API credentials are stored separately from source content and should never be returned to customers or model prompts in plaintext.
  • Redaction, confidence checks, and rate limits are defense layers. They do not replace your own security, privacy, or compliance review.