Build multi-step agent workflows
Coordinator→specialist agent graphs with approvals and traces.
One-shot Q&A is not enough for real workflows: triaging a bug report means searching docs, checking the changelog, and querying an internal API in sequence. Building that orchestration in-house means months of glue code, and you get no observability when it misbehaves.
BeforeQuery's Agent Platform lets you define coordinator→specialist agent graphs that combine retrieval, AI Actions, and MCP/A2A tools. Runs are fully traced for observability, risky steps route through a human-approval queue, and the @beforequery/agents SDK runs the same graphs from your own code.
How it works
- 1Define an agent graph: a coordinator that delegates to specialist agents.
- 2Give each specialist its own knowledge scope, AI Actions, and MCP/A2A tools.
- 3Insert human-approval nodes where a step is risky enough to need sign-off.
- 4Run graphs via the API or the @beforequery/agents SDK from your own code.
- 5Inspect full run traces — steps, retrievals, decisions — for every execution.
What you get
Features that make it work
Frequently asked questions
What can an agent in a graph actually do?
Specialists combine RAG retrieval over your indexed knowledge, AI Actions that call your APIs (with approval rules), and MCP/A2A tool calls to external systems. The coordinator routes the task across them.
How do I keep risky steps safe?
Human-approval nodes pause execution and queue the step for review; it resumes only after explicit approval. Approval queues, invocation logs, and run traces give you a full audit trail.
Can I run agents from my own backend?
Yes. The @beforequery/agents SDK provides runAgent and defineAgent, plus MCP and A2A clients, and there is a public run endpoint authenticated with client keys.
How long does setup take?
Most teams are live the same day. You connect a source — a docs URL, GitHub repo, Notion workspace, or a file upload — and BeforeQuery crawls, normalizes, chunks, and embeds the content automatically. The website widget is a single script tag, and the Slack/Discord bots install in a few clicks. There is no model to fine-tune and no infrastructure to run.
Are answers backed by citations?
Yes. Every answer includes a structured citations list with the URL, title, and excerpt of each source used, and the answer text cites sources inline. Users can click through to verify any claim against the original document — which is what makes the answers trustworthy enough for support, sales, and internal use.
Is my content used to train AI models?
No. Your indexed content, embeddings, and conversations are isolated to your workspace and are only used to answer your own questions. Content is never shared across customers, and data is encrypted at rest and in transit. Enterprise plans add OIDC/SAML SSO, audit logs, and data-residency options.
Which knowledge sources can I connect?
18+ source types: websites, GitHub, GitLab, Notion, Confluence, Google Drive, SharePoint, Shopify, Zendesk, Jira, Linear, Salesforce, Slack, Discourse, Stack Overflow, YouTube, OpenAPI specs, and direct file uploads (PDF, Markdown, TXT, HTML). Sources sync on a schedule, and GitHub/GitLab can re-index automatically via webhooks on every push.
What happens when the AI is not confident in an answer?
Answers are confidence-scored. When retrieval confidence falls below your threshold, BeforeQuery returns a transparent, configurable fallback — typically pointing users to your support channel — instead of guessing. Those unanswered questions are logged in knowledge gap analytics so you can close the gap in your docs.
Related use cases
Unblock developers integrating your API
Answer auth, webhook, and SDK questions with code-aware AI.
Put your docs inside the IDE via MCP
Cursor, Claude, and other MCP clients answer from your docs.
Fix developer portal search
Hybrid semantic + keyword search that actually finds the page.
Ready to build multi-step agent workflows?
Connect your knowledge sources and see cited AI answers in minutes — free, no credit card required.
Get Started Free