OutcomeCI

Official registry AI and agents

Create, validate, and manage agent-powered workflows with OutcomeCI.

Install

Add this to your client's MCP configuration (Claude Desktop: claude_desktop_config.json; Cursor: .cursor/mcp.json; VS Code and others use the same "mcpServers" shape).

Remote (streamable-http)
Claude Code
claude mcp add --transport http outcomeci 'https://api.outcomeci.com/v1/mcp'
JSON configuration (Claude Desktop, Cursor, VS Code)
{
    "mcpServers": {
        "outcomeci": {
            "type": "http",
            "url": "https://api.outcomeci.com/v1/mcp"
        }
    }
}

Documentation

From the project's README on GitHub, refreshed weekly. Links open on GitHub.

Contents
  1. OutcomeCI MCP Server
  2. Connect
  3. Try it
  4. Tools
  5. Permissions and credentials
  6. Pricing
  7. Documentation and feedback
  8. Repository metadata
  9. License

OutcomeCI MCP Server

Connect your AI assistant to OutcomeCI to create, validate, and manage agent-powered workflows across the tools your team uses.

This is the public connection and documentation repository for OutcomeCI's hosted Model Context Protocol (MCP) server. It contains server metadata and client configuration—not the hosted backend or a self-hosted server distribution.

Connect

Add OutcomeCI as a remote MCP server in your client:

https://api.outcomeci.com/v1/mcp
  • Transport: Streamable HTTP
  • Authentication: OAuth through your browser
  • Access: Sign in to OutcomeCI and authorize a workspace

No local server installation or shared API key is required. Each person authorizes their own connection. Your client must support remote Streamable HTTP servers and OAuth.

Setup guides

For clients that accept this configuration format, see the example configuration. Other clients may use different field names; follow their remote-server setup instructions.

Try it

Ask your connected assistant:

List the workflows in my OutcomeCI workspace.

Read the workflow schema and help me create a workflow that turns a Slack request into a GitHub pull request. Validate it before saving.

Show the recent runs for my workflow and inspect the logs for the latest failure.

Available tools depend on the permissions granted to the connection. Configuring a workflow may also require connectors, credentials, and an agent connection.

Tools

The server exposes the following tools. Your client's live tools/list response is authoritative for the current schemas and authorized tool set.

ToolWhat it does
list_workspacesShow the workspace authorized for the connection.
list_workflowsList workflows and their latest metadata.
get_workflowRead a workflow's latest definition and support files.
get_workflow_schemaGet the current outcome.yml contract and an example.
get_connector_schemaInspect a connector's operations and credential requirements.
get_integration_setup_guideGet setup guidance for an integration.
validate_workflowValidate a complete workflow package without saving it.
create_workflowCreate a workflow and its first version.
upload_workflow_versionSave a new immutable version with checks against concurrent changes.
configure_workflow_triggerConfigure supported email or webhook triggers.
create_credential_depositCreate a secure UI link for depositing a credential.
get_credential_deposit_statusCheck a deposit's status without retrieving its secret.
list_vault_entriesInspect credential metadata and grants, not secret values.
list_agent_connectionsList connected agents' metadata.
list_workflow_runsList recent runs for a workflow.
get_workflow_run_logsRead paginated run logs.
cancel_workflow_runRequest cancellation of an active run.

Permissions and credentials

Authorize only the workspace and permissions you intend to give your assistant. Use your client's approval controls for changes such as saving workflows, configuring triggers, or cancelling runs.

Never paste credentials into chat. Use the secure deposit link when a workflow needs a secret. Treat deposit links, inbound webhook URLs, and run logs as potentially sensitive; do not include them in public issues.

Pricing

OutcomeCI Cloud usage is billed according to OutcomeCI pricing. Connecting an assistant does not make workflow execution free.

Documentation and feedback

For documentation problems or feature requests, open an issue. Include the client name and a sanitized description; never include tokens, secrets, or private workflow data.

Repository metadata

server.json describes the hosted endpoint using the MCP Registry server schema. Its version tracks this repository's metadata, not a pinned version of the hosted service. Including the manifest does not imply that a directory or registry has approved the listing.

License

This repository's documentation, metadata, and configuration examples are licensed under the MIT License. The hosted backend is not distributed in this repository.

Frequently asked questions

What is OutcomeCI?

Create, validate, and manage agent-powered workflows with OutcomeCI.

How do I install OutcomeCI?

It runs as a remote server. In Claude Code: claude mcp add --transport http outcomeci 'https://api.outcomeci.com/v1/mcp'. For other clients, paste the JSON configuration from this page into their MCP settings.

Does OutcomeCI need an API key?

Its registry entry lists no required key or secret. Some features may still need an account with the service it connects to.

Is OutcomeCI open source?

Yes. Its code is on GitHub under the MIT license.

Who maintains OutcomeCI?

It is published by outcomeci. AI.tn lists it but does not maintain or endorse it.

From the official MCP Registry. Listing is not an endorsement: review a server's code and permissions before giving it access to your data.