Actera
Review AI agent decisions and approvals and draft, test and publish guardrail policies
- Platforms
- API, Web
- Runs
- Remotely
- Version
- 1.0.0
- Listing updated
- 5 days ago
Similar MCP servers
About Actera
Actera is an MCP server for security. It runs as a hosted remote server, so there is nothing to install locally.
Actera sits between an agent and the systems it acts on. An integration asks Actera before acting, and Actera checks the request against the agent's assigned policies and returns allow, require approval or block. Policies can depend on the action, trusted business context and cumulative usage, so several small refunds can still hit a daily allowance.
This server gives an assistant access to the control side: review agent decisions and approvals, then draft, test and publish policies. A suggested path is to connect one agent and operation, study real requests in Decisions, build a focused policy, test expected outcomes, publish and assign it, then switch the integration from monitoring to enforcing. Assigning policies alone does not enforce them.
Platform and security teams define the boundaries, application teams enforce them where actions are dispatched, and reviewers handle exceptions. The docs warn that a decision does not prove a downstream operation completed and that the application must place the check where it can still stop the action.
Install
Run in a terminal
claude mcp add --transport http actera https://mcp.actera.dev/mcpAdd to Settings → Connectors → Add custom connector
https://mcp.actera.dev/mcpAdd to ~/.cursor/mcp.json
{
"mcpServers": {
"actera": {
"url": "https://mcp.actera.dev/mcp"
}
}
}Add to .vscode/mcp.json
{
"servers": {
"actera": {
"type": "http",
"url": "https://mcp.actera.dev/mcp"
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"actera": {
"serverUrl": "https://mcp.actera.dev/mcp"
}
}
}Add to Grok Bot desktop app → Connect apps → Add a custom connector
https://mcp.actera.dev/mcpEnter a Name (for example "Actera"), choose Type: Remote HTTPS, paste this as the Server URL and choose Add plugin.
Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"actera": {
"type": "http",
"url": "https://mcp.actera.dev/mcp"
}
}
}Using several servers? Build one config file with the MCP config generator.
What it does
- Decision review
- Browse what agents asked to do and how Actera ruled.
- Approval handling
- See requests waiting for an eligible reviewer.
- Policy drafting
- Write policies based on real requests.
- Policy testing
- Check expected outcomes before publishing.
- Cumulative limits
- Policies can consider usage across requests.
- Monitor then enforce
- Run in monitoring mode before enforcement.
Good for
- Turn real agent requests into a focused policy
- Test a refund limit before publishing it
- Review pending approval requests
- Audit why an agent action was blocked
Actera FAQ
What decisions can Actera return?
Allow, require approval or block.
Does assigning a policy enforce it?
No. Enforcement must also be enabled in the plugin or application.
Does a decision prove the action completed?
No, the docs say it does not independently prove that.
Which setups are documented?
TypeScript quickstart, Claude Code, Codex and GitHub Copilot.
How do I add Actera to Claude Code?
Run this command in your terminal: claude mcp add --transport http actera https://mcp.actera.dev/mcp
What are the best alternatives to Actera?
The closest alternatives to Actera are Marrow AI: Control Layer for AI Agents, Abe, Forge AI Accountability, Antaeus and Ayedesk.
Where this comes from
- Description: written from the project's website.
- Install steps: from the official MCP Registry entry.
