# Devctrl

## Docs

- [Audit Log](https://docs.devctrl.ai/concepts/audit-log.md): Every tool call is logged — who requested it, what was decided, and why. Full transparency for compliance.
- [Identities](https://docs.devctrl.ai/concepts/identities.md): An identity represents an AI agent in Devctrl. It carries credentials, labels, and optional tool restrictions.
- [MCP Gateway](https://docs.devctrl.ai/concepts/mcp-gateway.md): The managed gateway that routes agent requests to upstream tool servers through policy checks.
- [Policies](https://docs.devctrl.ai/concepts/policies.md): Define rules that control what your agents can do. Policies are CEL expressions evaluated on every tool call.
- [Task-Based Access Control](https://docs.devctrl.ai/concepts/task-based-access-control.md): Grant agents only the permissions their current task requires — scoped, time-limited, and auditable.
- [Tasks](https://docs.devctrl.ai/concepts/tasks.md): Tasks define what an agent is trying to do. They carry context, scope permissions, and expire automatically.
- [Create Identities](https://docs.devctrl.ai/console/create-identities.md): Create agent identities with credentials, labels, and optional tool restrictions.
- [Define Tasks](https://docs.devctrl.ai/console/define-tasks.md): Create task definitions with context schemas and tool restrictions for scoped access.
- [Manage Tools](https://docs.devctrl.ai/console/manage-tools.md): View and configure the tools available in your project, discovered from registered MCP servers.
- [Projects](https://docs.devctrl.ai/console/projects.md): Create and manage projects — the top-level container for all your agents, tools, and policies.
- [Publish Policies](https://docs.devctrl.ai/console/publish-policies.md): Create versioned policy releases for the gateway to enforce.
- [Register MCP Servers](https://docs.devctrl.ai/console/register-mcp-servers.md): Connect upstream MCP tool servers so agents can discover and call tools through the gateway.
- [View Executions](https://docs.devctrl.ai/console/view-executions.md): Browse the audit log to see every tool call, policy decision, and agent action.
- [Write Policies](https://docs.devctrl.ai/console/write-policies.md): Author CEL-based policy rules that control what your agents can do.
- [Connect an Agent](https://docs.devctrl.ai/gateway/connect-an-agent.md): Point any MCP-compatible AI agent at the Devctrl gateway.
- [Gateway Overview](https://docs.devctrl.ai/gateway/overview.md): The Devctrl gateway authenticates agents, enforces policies, and logs every tool call.
- [Task Sessions](https://docs.devctrl.ai/gateway/task-sessions.md): Create scoped, time-limited tokens that bind an agent to a specific task with context.
- [How It Works](https://docs.devctrl.ai/how-it-works.md): The Devctrl architecture — how agents, the gateway, policies, and your tools connect.
- [Introduction](https://docs.devctrl.ai/introduction.md): Devctrl is a policy and governance layer for AI agents. Control what your agents can do, when they can do it, and audit every action.
- [CEL Expressions](https://docs.devctrl.ai/policies/cel-expressions.md): Reference guide for writing CEL policy expressions in Devctrl.
- [Policy Examples](https://docs.devctrl.ai/policies/examples.md): Real-world policy patterns for common access control scenarios.
- [RBAC vs TBAC](https://docs.devctrl.ai/policies/rbac-vs-tbac.md): Why traditional role-based access control doesn't work for AI agents, and how Task-Based Access Control solves it.
- [Quickstart](https://docs.devctrl.ai/quickstart.md): Go from zero to your first policy-enforced tool call in 15 minutes.