Agents
Lists agents in the registry with status, runtime, model, and governance profile. Search or filter, then open an agent to manage it.
| Name | Agent ID | Status | Runtime | Model | Updated |
|---|
Create Agent
Creates an agent. Governance is optional. Optionally map it to a project.
Agent Details
Shows this agent’s configuration, governance profile, lifecycle actions, and feedback history.
—
Edit Agent
Edits this agent’s name, runtime, tools, data sources, and governance profile. Use Agent Details for activate, disable, or archive.
—
Query Tester
Runs a live execute call on an active agent. Shows the response, token usage, governance decision, and a form to submit feedback.
Execution result
Answer
—
Run details
Raw response JSON
Governance JSON
Governance
Create and review governance profiles that agents use at execute time. Each profile holds who can call (roles), rate limits, exact pattern redaction/blocks, and custom policy checks evaluated for meaning (not only exact words). Attach a profile on Create or Edit Agent.
Profile catalog
Profiles already available to attach to agents, including which agents use each one and their custom policy checks.
Decision View
Compact audit log of governance checks. Each execute creates two checks by design — before the model (can this caller run?) and after (is the response allowed). Click a row for details.
Projects
Projects are parent containers for agents. After create you can
Publish (give others your endpoint + fab_… key) and
Connect (attach another Fabric project or external API/MCP with their credentials).
Project registry
Project
Lifecycle
Create → Test → Activate → Publish → Deprecate. Test runs a smoke execute; Activate enables execute; Publish marks the project published.
Teams
Teams under this project. Each team can have its own governance profile.
Publish — let others use your agents
Issue endpoints + fab_… keys so other projects/apps can call agents mapped here.
Send Authorization: Bearer fab_…. Atlas stays behind Fabric.
Example curl (API publish)
Connect — attach other projects / sources
Inbound: store their credentials so this project can reach another Fabric project or an external API/MCP.
Mapped agents
Answer
—
Raw result JSON
Knowledge · TwymAtlas
TwymAtlas is Fabric’s only knowledge plane (MCP). Fabric does not host generic or local knowledge sources — agents retrieve indexed enterprise knowledge from Atlas at execute time.
Connection status
Answer
How to use Atlas from Fabric
- Start TwymAtlas and open its MCP page (status must be Active).
- Copy the MCP URL and API key exactly from that page into Fabric
.env(ATLAS_MCP_URL,ATLAS_API_KEY), then restart Fabric. - If Atlas UI shows a different origin (e.g.
:8080), either open Atlas on:8000or setATLAS_MCP_URLto that origin’s/mcp/URL. - After Reconfigure / Revoke on Atlas MCP, you must paste the new
tf_…key into Fabric.env— old keys return 401. - Use a knowledge agent (seeded:
agt_ops_knowledge) or any agent with toolknowledge_search. - Via project: open Twym Ops → map a knowledge agent → Execute (pass agentId if multiple agents) with a message like “search knowledge for onboarding policy”.
- Direct test: Query Tester on Ops Knowledge Specialist.
- If Atlas returned chunks, the execute response includes
citations/knowledgeUsed: true.
Route not found
That AI Services path is not configured.