Docs
TraceMem documentation
For the platform, security and architecture teams checking how TraceMem fits their environment: how it deploys into your network, what it connects to, and how your agents reach it.
- Runs
- Inside your network
- Identity
- Your OIDC provider
- Connects to
- 84 named systems, plus any API or MCP server
- Agents connect over
- MCP, HTTP and our SDKs
Works with the stack you already run.
Your identity provider, your data systems, your models and the tools your approvers already work in.
- Microsoft Entra ID
- Okta
- Salesforce
- Snowflake
- PostgreSQL
- SAP S/4HANA
- Slack
- Microsoft Teams
- Anthropic
- OpenAI
- Amazon Bedrock
- and 73 moreplus any API or MCP server
Will it run in our environment?
Deployment
How TraceMem is delivered, where it sits in your stack, and what connects to it.
- How it fits your stack
- Distributions
- Topologies
- What connects to TraceMem
- What TraceMem connects to
- Your environment
- Security and operations
How do our agents connect to it?
SDKs and APIs
How agents and applications reach TraceMem, and how to install the SDKs and plugins.
- The surfaces
- Install an SDK
- What the SDKs cover
- Agent MCP server
- HTTP runtime API
- Model gateway endpoints
- Admin MCP server
- Agent frameworks and coding agents
Does it work with our systems?
Integrations
The identity providers, data sources, approval channels, models and agent platforms it connects to.
- Identity providers
- Data sources
- Approval channels
- Models
- Agent platforms
- Logs, audit and export
Everything to evaluate it, and to run it.
These pages cover what a technical evaluation needs. Customers get the rest through the customer portal, and a demo with our team can walk through how TraceMem fits your own stack.
- On this site
- How it deploys, the distributions and topologies, every integration, and how to install the SDKs and plugins.
- In the customer portal
- Installation guides, the configuration and API references, downloads and licence keys.