Rezoomex MCP Server

Enterprise contract-based development environment powered by the Model Context Protocol. Access your project context, user stories, and contract specifications directly in your development tools.

Get Started Learn About MCP

How It Works

Connect your IDE to Rezoomex in three simple steps

1

Add MCP Server

Copy the configuration for your IDE and add it to your MCP settings file

2

Authenticate in IDE

Open your IDE's MCP settings page and click the authentication prompt for the Rezoomex server. This will open your browser to sign in with your Rezoomex credentials

3

Access Your Playbooks

Use MCP tools to fetch product info, user stories, personas, and more—all scoped to your contracts

Understanding the Authentication Flow

🔐 Add MCP Server Configuration

After adding the MCP server configuration to your IDE, check your IDE's MCP settings page. You'll see the Rezoomex MCP server listed.

🌐 Authentication Prompt

Your IDE will prompt you to authenticate. Click the authentication link/button in your IDE's MCP settings. This will open your browser and redirect you to Auth0 to sign in with your Rezoomex credentials. If your browser does not open on its own, the IDE shows the sign-in URL — copy the link and open it manually in any browser.

✅ Authorization Complete

After successful login in your browser, you'll be redirected back. You can close the browser tab at that point — return to your IDE and verify the connection status shows as authenticated.

🔄 Using MCP Tools

Once authenticated, you can use any MCP tool (fetchProductInfo, getUserInfo, etc.). The IDE stores your authentication token securely for future sessions.

Model Context Protocol Integration

The MCP Server provides smart contractors with precise access to project scope and context. Built on the Model Context Protocol, it delivers exactly what you need for your contracted work—nothing more, nothing less.

Whether you're using GitHub Copilot in VS Code, Cursor, Claude, or any MCP-compatible IDE, the server connects your development environment to the latest project context, securely and reliably.

Product Playbook Access

Every contract is tied to a comprehensive Product Playbook containing all necessary resources:

  • User stories and requirements
  • Test cases and validation criteria
  • User personas and context
  • Acceptance criteria and milestones

Access this context directly in your development environment to stay aligned with contract goals without chasing documentation or waiting for clarifications.

Secure Access Control

Enterprise-grade OAuth authentication with contract-scoped access control ensures security and precision:

  • Access only projects and resources linked to your active contracts
  • Context and code access filtered based on agreement specifications
  • Controlled exposure with clear boundaries for confident delivery

Enterprise Benefits

  • Accelerated development with structured, contextual inputs
  • Security-first approach with contract-scoped access
  • Real-time synchronization with evolving project requirements
  • Direct integration with contract specifications and deliverables

Available Tools

The MCP server provides the following tools to access your contract-scoped project data:

  • getUserInfo: Get your current user information and access permissions
  • checkNdaStatus: Verify your NDA status for accessing protected resources
  • fetchProductInfo: Access comprehensive product playbook information including name, description, and project environment
  • fetchElevatorPitch: Retrieve the elevator pitch for any product in your scope
  • fetchVisionStatement: Get the vision statement for strategic context
  • fetchProjectEnvironment: Access technical stack information (tools, frameworks, languages, services)
  • listPersonas: List all personas defined for a project
  • fetchPersona: Get detailed persona information including demographics, motivations, and pain points
  • fetchUserStoriesForPersona: List all user stories associated with a specific persona (with pagination)
  • fetchJtbdForPersona: Retrieve Jobs To Be Done items describing a persona's goals and motivations
  • fetchEventsForPersona: Get the user journey event sequence for a specific persona
  • fetchUserStory: Retrieve a user story with actor, goal, and description
  • fetchUserStoryAcceptanceCriteria: Get Given/When/Then acceptance criteria for a user story
  • fetchUserStoryWireframes: Access wireframe images and descriptions for a user story
  • fetchTestCasesForUserStory: Retrieve numbered test cases with steps for a user story
  • fetchTestDataForUserStory: Get test data tables for a user story
  • fetchHTML: Fetch raw HTML from a wireframe
  • fetchCSS: Fetch CSS styles from a wireframe
  • fetchArchitecturalNotes: Access architectural documentation with table-of-contents or per-section retrieval
  • fetchERDiagram: Get the Entity Relationship Diagram — tables, columns, and foreign-key relationships
  • updateTestCaseStatus: Record a PASSED or FAILED result with a comment for a test case

Getting Started

  1. Receive MCP endpoint: Access credentials provided via email or Rezoomex dashboard
  2. Authenticate via OAuth: Secure login using your Rezoomex credentials
  3. Access project context: View and integrate contract-linked playbook data
  4. Begin development: Work with your preferred tools and scoped project requirements

Support

Reference the integration guide below or contact your project administrator for assistance with setup and configuration.

Available MCP Tools

Access your contract-scoped project data through these tools

getUserInfo

User Info

Get your current user information and access permissions

View Example

Usage: No parameters required

Returns: User ID, email, and client information

When to use: Verify your authentication and check what contracts you have access to

fetchProductInfo

Product

Access comprehensive product playbook information

View Example

Parameters:

  • projectId: The project identifier (e.g. CJTY) — productCode is still accepted as a legacy alias for the same value

Returns: Name, description, elevator pitch, vision statement, and project environment

When to use: Get a comprehensive overview of the product you're working on

fetchUserStory

User Stories

Retrieve core user story details — actor, goal, and description — with guidance on companion tools for full details

View Example

Parameters:

  • projectId: The project identifier (e.g. CJTY)
  • parentId: The persona ID (e.g. CJTY-P-001)
  • storyId: The user story ID (e.g. CJTY-US-001)

Returns: Core story details (actor, goal, description) plus next-steps hints pointing to acceptance criteria, wireframes, and test case tools

When to use: Start here — the response tells you exactly which companion tools to call next

fetchUserStoryAcceptanceCriteria

User Stories

Retrieve acceptance criteria and Given/When/Then scenarios for a user story

View Example

Parameters:

  • projectId, storyId

Returns: Human-readable text plus a structuredContent object with one normalized item per saved resource (resourceId, parentId, resourceType, properties) — records are never merged, so each saved criterion keeps its own stable ID. A non-empty saved record with a missing or duplicate resourceId is reported as an error (isError: true) instead of being silently included.

When to use: Understand what "done" looks like for a story before writing code, or to safely target a specific saved criterion for update/removal by its resourceId

fetchUserStoryWireframes

User Stories

Retrieve wireframe images and descriptions for a user story

View Example

Parameters:

  • projectId, storyId

Returns: Wireframe URLs and layout descriptions, each tagged with its resource ID when available

When to use: Understand the expected UI before building a screen

fetchTestCasesForUserStory

User Stories

Retrieve all test cases for a user story, formatted as numbered headings with a sequential implementation tip

View Example

Parameters:

  • projectId, storyId

Returns: Numbered test cases with titles, steps, expected results, priorities, resource IDs (when available), and a tip to implement one at a time

When to use: Review test requirements before or after implementation — implement each test case sequentially

fetchTestDataForUserStory

User Stories

Retrieve test data datasets for a user story, each formatted as a named markdown table with column headers and data rows

View Example

Parameters:

  • projectId, storyId

Returns: Named datasets with column headers and CSV rows rendered as markdown tables, each tagged with its resource ID when available

When to use: Understand what test data exists for a story before writing or running tests

fetchHTML

Technical

Extract raw HTML source code from a story's wireframe

View Example

Parameters:

  • projectId, personaId, storyId

Returns: HTML markup ready to use as a starting point

When to use: Bootstrap a screen from the wireframe's HTML — use fetchUserStoryWireframes first to confirm HTML exists

fetchCSS

Technical

Extract CSS styles from a story's wireframe

View Example

Parameters:

  • projectId, personaId, storyId

Returns: CSS stylesheet for the wireframe

When to use: Get baseline styles to match the design spec — use fetchUserStoryWireframes first to confirm CSS exists

fetchPersona

Personas

Get detailed persona information

View Example

Parameters:

  • projectId: The project identifier (e.g. CJTY)
  • personaId: The persona identifier

Returns: Human-readable text (demographics, motivations, pain points, goals) plus a structuredContent object: { status, projectId, resourceId, parentId, resourceType, properties, isComplete }. properties is every saved field passed through verbatim — including empty strings, empty arrays, nulls, and fields the text view does not render — so the exact saved values are available if a persona is updated out of band (this server exposes no persona write tool). A response that is missing resourceId, properties, parentId or resourceType, or whose identity does not match what was requested, is reported as an error (isError: true, status: "error", error: "MALFORMED_RESPONSE") instead of a partial result.

When to use: Understand who you're building for, or read the exact saved persona fields from structuredContent.properties when a persona has been changed outside this server

listPersonas

Personas

List all personas for a project

View Example

Parameters: projectId (required)

Returns: Human-readable text plus a structuredContent object: { status, projectId, items, itemCount, isComplete } where each item is { resourceId, name, role } (name/role are omitted when not saved). A project with no personas returns items: [], itemCount: 0, isComplete: true and is not an error. A persona missing its resourceId is reported as an error (isError: true, error: "MALFORMED_RESPONSE") with no partial items.

When to use: Discover all user personas defined for a project, then pass an item's resourceId to fetchPersona

fetchJtbdForPersona

JTBD

Fetch Jobs To Be Done (JTBD) items for a specific persona

View Details

Parameters: projectId (required), personaId (required), pageSize (optional), startOffset (optional)

When to use: After identifying a persona with listPersonas, fetch their JTBD items to understand goals and motivations.

fetchEventsForPersona

User Journey

Fetch events (user journey / interaction sequence) for a specific persona

View Details

Parameters: projectId (required), personaId (required), pageSize (optional), startOffset (optional)

When to use: After identifying a persona with listPersonas, fetch their user journey events to understand the sequence of interactions between the user and the system, including triggers and system responses.

fetchElevatorPitch

Product

Retrieve the product elevator pitch

View Example

Parameters: projectId

When to use: Quick product summary for context

fetchVisionStatement

Product

Get the vision statement for strategic context

View Example

Parameters: projectId

When to use: Align work with product vision

fetchProjectEnvironment

Technical

Access technical stack information

View Example

Returns: Tools, frameworks, languages, services, and dependencies

When to use: Understand the tech stack for the project

fetchArchitecturalNotes

Technical

Fetch architectural notes for a project. Call without a section to get a table of contents, then specify a section name to retrieve full details for that section.

View Example

Parameters:

  • projectId: The project identifier (e.g. K644)
  • section (optional): One of introduction, howItWorks, systemOverview, systemComponents, architecturalGoals, securityArchitecture, highLevelArchitecture, deploymentArchitecture, technologyStackSummary

Returns: When section is omitted — a table of contents listing all available sections with item counts. When section is specified — full markdown detail for that section (prose, numbered lists, or tables as appropriate).

When to use: Start without a section to discover what's documented, then fetch individual sections to understand system architecture, technology choices, security approach, or deployment strategy.

fetchLeanCanvas

Product

Fetch the Lean Canvas for a project — problem, solution, unique value proposition, unfair advantage, customer segments, channels, key metrics, cost structure, and revenue streams.

View Example

Parameters: projectId

When to use: Understand or validate the business model behind a product

fetchERDiagram

Technical

Fetch the Entity Relationship Diagram (ERD) for a project. Returns all database tables with their columns and the foreign-key relationships between them.

View Example

Parameters:

  • projectId: The project identifier (e.g. 7UL4)

Returns: All tables with column names, types, and PK/FK flags, followed by a relationship table showing foreign-key links between tables.

When to use: Use when you need to understand the data model — table structure, column types, or how entities relate to each other via foreign keys.

fetchUserStoriesForPersona

User Stories

List all user stories for a specific persona

View Example

Parameters: projectId, personaId, pagination options

Returns: A valid persona with zero saved stories is a normal successful result (not an error). Genuine failures — persona not found, network/auth errors — set isError: true on the response so callers can distinguish "no stories yet" from "something went wrong" without parsing text.

When to use: See all stories for a user segment

checkNdaStatus

Access

Verify NDA status for protected resources

View Example

When to use: Check if you have access to confidential project data

updateTestCaseStatus

QA

Update the status (PASSED or FAILED) and add a comment for a test case. Requires an active QA smart contract for the story.

View Example

Parameters:

  • projectId (required): Project ID, e.g. R27A
  • storyId (required): Story ID, e.g. R27A-US-034
  • testCaseId (required): Test case resource ID, e.g. R27A-TC-001
  • status (required): Test result status — PASSED or FAILED
  • comment (required): Comment explaining the test result

Returns: Confirmation of the updated test case status with the recorded comment

When to use: After running a test case, record the outcome. Use fetchTestCasesForUserStory first to get valid test case IDs.

fetchPlaybookTree

Owner

Returns the whole project skeleton in one call — every saved resource id, its type and its parent, without the resource contents. Also reports resources whose parent has been deleted, which per-parent lookups cannot surface. Project owners only.

View Example

Parameters:

  • projectId (required): Project ID, e.g. R27A

Returns: tree (resources nested under their parent), orphans (resources whose parent no longer exists) and cycles, as ids, types and parents only

When to use: Before creating, updating or removing playbook resources, to see what already exists — one call instead of listing each parent in turn. Always re-fetch rather than reusing a tree from earlier in the conversation, which may be stale.

getLastUsedIds

Owner

Reports the highest id number currently in use for each resource type in the project. Numbers come from resources that still exist, so a number may have gaps below it. Project owners only.

View Example

Parameters:

  • projectId (required): Project ID, e.g. R27A

Returns: One entry per resource type with resourceType, lastUsedNumber and lastUsedId

When to use: When minting the next id for a resource type. Note the highest id of a deleted resource is not tracked, so this reports the highest surviving id rather than a high-water mark.

Agentic Playbook Tools (dev / qa only) — the 37 tools below are hidden by default. A client must send x-rezoomex-playbook-type: agentic against /dev/mcp or /qa/mcp to see them; they never appear on the production route (/mcp) regardless of header. Underlying API response shapes for these resource types aren't fully confirmed yet, so results are shown as generic JSON.

fetchAgentArchitecture

Agentic

Fetch the agent architecture design for a project — agent roles and orchestration design. Agentic-playbook resource, available only on dev/qa with playbook type set to agentic.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Agent Architecture data as generic JSON (resourceId, resourceType, properties)

fetchMemory

Agentic

Fetch the memory design for a project — what the agent remembers and how memory is structured.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Memory Design data as generic JSON (resourceId, resourceType, properties)

fetchSession

Agentic

Fetch the session design for a project — session lifecycle and state management.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Session Design data as generic JSON (resourceId, resourceType, properties)

fetchCI_Premerge

Agentic

Fetch the pre-merge CI pipeline configuration and checks for a project.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Pre-Merge CI Pipeline data as generic JSON (resourceId, resourceType, properties)

fetchCI_Postmerge

Agentic

Fetch the post-merge CI pipeline configuration and checks for a project.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Post-Merge CI Pipeline data as generic JSON (resourceId, resourceType, properties)

fetchCI_Production

Agentic

Fetch the production CI/deployment pipeline configuration for a project.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Production CI Pipeline data as generic JSON (resourceId, resourceType, properties)

fetchObservabilityDashboard

Agentic

Fetch the observability dashboard configuration for a project.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Observability Dashboard data as generic JSON (resourceId, resourceType, properties)

fetchEvalDashboard

Agentic

Fetch the evaluation dashboard configuration for a project.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Eval Dashboard data as generic JSON (resourceId, resourceType, properties)

fetchHumanReview

Agentic

Fetch the trace currently escalated for human review for a project, if any. At most one trace is under active review at a time, so this has no separate list tool.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Human Review data as generic JSON (resourceId, resourceType, properties)

listToolsRegistry

Agentic

List the tools registry for a project (one per project). Returns its resourceId (toolsRegistryId), needed by listTools and fetchTool.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Tools Registry data as generic JSON (resourceId, resourceType, properties)

listContextAssembly

Agentic

List the context assembly for a project (one per project). Returns its resourceId (contextAssemblyId), needed by listStages and fetchStage.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Context Assembly data as generic JSON (resourceId, resourceType, properties)

listEvalRubric

Agentic

List the eval rubric for a project (one per project). Returns its resourceId (evalRubricId), needed by listRubrics and fetchRubric.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Eval Rubric data as generic JSON (resourceId, resourceType, properties)

listGoldenDataset

Agentic

List the golden dataset for a project (one per project). Returns its resourceId (goldenDatasetId), needed by listScenarios and fetchScenario.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Golden Dataset data as generic JSON (resourceId, resourceType, properties)

listSecurityThreatModel

Agentic

List the security threat model for a project (one per project). Returns its resourceId (securityThreatModelId), needed by listThreats and fetchThreat.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Security Threat Model data as generic JSON (resourceId, resourceType, properties)

listObservabilityMetrics

Agentic

List the observability metrics collection for a project (one per project). Returns its resourceId (observabilityMetricsId), needed by listMetrics and fetchMetric.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Observability Metrics data as generic JSON (resourceId, resourceType, properties)

listObservabilityLogs

Agentic

List the observability logs collection for a project (one per project). Returns its resourceId (observabilityLogsId), needed by listLogs and fetchLog.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Observability Logs data as generic JSON (resourceId, resourceType, properties)

listFailureModes

Agentic

List the failure modes collection for a project (one per project). Returns its resourceId (failureModesId), needed by listModes and fetchFailureMode.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: Failure Modes data as generic JSON (resourceId, resourceType, properties)

listLLMJudgement

Agentic

List the LLM judgement for a project (one per project). Returns its resourceId (llmJudgementId), needed by listTraces and fetchTraceForLLMJudgement.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY

Returns: LLM Judgement data as generic JSON (resourceId, resourceType, properties)

listTools

Agentic

List the tools available in a project's tools registry. Get the toolsRegistryId from listToolsRegistry. Use fetchTool with a toolId from this list to get full detail on one.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • toolsRegistryId (required): The Tools Registry ID, e.g. CJTY-TR-001. Get it from listToolsRegistry

Returns: Tools Registry Tools data as generic JSON (resourceId, resourceType, properties)

listTraces

Agentic

List execution traces for a project, each with its judgement outcome. Use fetchTraceForLLMJudgement or fetchObservabilityTrace with a traceId from this list.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • llmJudgementId (required): The LLM Judgement ID, e.g. CJTY-LJ-001. Get it from listLLMJudgement

Returns: Traces data as generic JSON (resourceId, resourceType, properties)

listModes

Agentic

List the failure modes in a project's failure modes collection. Use fetchFailureMode with a failureModeId from this list.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • failureModesId (required): The Failure Modes ID, e.g. CJTY-FM-001. Get it from listFailureModes

Returns: Failure Mode Entries data as generic JSON (resourceId, resourceType, properties)

listMetrics

Agentic

List observability metrics tracked for a project. Use fetchMetric with a metricId from this list.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • observabilityMetricsId (required): The Observability Metrics ID, e.g. CJTY-OM-001. Get it from listObservabilityMetrics

Returns: Observability Metrics data as generic JSON (resourceId, resourceType, properties)

listLogs

Agentic

List observability log entries for a project. Use fetchLog with a logId from this list.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • observabilityLogsId (required): The Observability Logs ID, e.g. CJTY-OL-001. Get it from listObservabilityLogs

Returns: Observability Logs data as generic JSON (resourceId, resourceType, properties)

listScenarios

Agentic

List test scenarios in a project's golden dataset. Use fetchScenario with a scenarioId from this list.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • goldenDatasetId (required): The Golden Dataset ID, e.g. CJTY-GD-001. Get it from listGoldenDataset

Returns: Golden Dataset Scenarios data as generic JSON (resourceId, resourceType, properties)

listThreats

Agentic

List threats identified in a project's security threat model. Use fetchThreat with a threatId from this list.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • securityThreatModelId (required): The Security Threat Model ID, e.g. CJTY-ST-001. Get it from listSecurityThreatModel

Returns: Security Threats data as generic JSON (resourceId, resourceType, properties)

listStages

Agentic

List stages in a project's context assembly pipeline. Get the contextAssemblyId from listContextAssembly. Use fetchStage with a stageId from this list.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • contextAssemblyId (required): The Context Assembly ID, e.g. CJTY-CA-001. Get it from listContextAssembly

Returns: Context Assembly Stages data as generic JSON (resourceId, resourceType, properties)

listRubrics

Agentic

List evaluation rubrics defined for a project. Use fetchRubric with a rubricId from this list.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • evalRubricId (required): The Eval Rubric ID, e.g. CJTY-ER-001. Get it from listEvalRubric

Returns: Eval Rubrics data as generic JSON (resourceId, resourceType, properties)

fetchTool

Agentic

Fetch the full spec/definition for one tool from the tools registry. Get the toolId from listTools.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • toolsRegistryId (required): The Tools Registry ID, e.g. CJTY-TR-001. Get it from listToolsRegistry
  • toolId (required): The toolId (e.g. 'CJTY-TL-001')

Returns: Tool data as generic JSON (resourceId, resourceType, properties)

fetchTraceForLLMJudgement

Agentic

Fetch the LLM-as-judge verdict and reasoning for one execution trace. Get the traceId from listTraces.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • llmJudgementId (required): The LLM Judgement ID, e.g. CJTY-LJ-001. Get it from listLLMJudgement
  • traceId (required): The traceId (e.g. 'CJTY-TRACE-001')

Returns: LLM Judgement data as generic JSON (resourceId, resourceType, properties)

fetchObservabilityTrace

Agentic

Fetch raw execution/observability detail (spans, timing, logs) recorded under one trace; the traceId is the parent trace. Get the traceId from listTraces. This is a different view of the same trace than fetchTraceForLLMJudgement.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • traceId (required): The traceId (e.g. 'CJTY-TRACE-001')

Returns: Observability Trace data as generic JSON (resourceId, resourceType, properties)

fetchFailureMode

Agentic

Fetch detail for one specific failure mode. Get the failureModeId from listModes.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • failureModesId (required): The Failure Modes ID, e.g. CJTY-FM-001. Get it from listFailureModes
  • failureModeId (required): The failureModeId of the individual mode, not the collection (e.g. 'CJTY-MODE-001')

Returns: Failure Mode data as generic JSON (resourceId, resourceType, properties)

fetchMetric

Agentic

Fetch detail and values for one specific observability metric. Get the metricId from listMetrics.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • observabilityMetricsId (required): The Observability Metrics ID, e.g. CJTY-OM-001. Get it from listObservabilityMetrics
  • metricId (required): The metricId (e.g. 'CJTY-METRIC-001')

Returns: Metric data as generic JSON (resourceId, resourceType, properties)

fetchLog

Agentic

Fetch one specific observability log entry. Get the logId from listLogs.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • observabilityLogsId (required): The Observability Logs ID, e.g. CJTY-OL-001. Get it from listObservabilityLogs
  • logId (required): The logId (e.g. 'CJTY-LOG-001')

Returns: Log Entry data as generic JSON (resourceId, resourceType, properties)

fetchScenario

Agentic

Fetch one test scenario from a project's golden dataset. Get the scenarioId from listScenarios.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • goldenDatasetId (required): The Golden Dataset ID, e.g. CJTY-GD-001. Get it from listGoldenDataset
  • scenarioId (required): The scenarioId (e.g. 'CJTY-SC-001')

Returns: Scenario data as generic JSON (resourceId, resourceType, properties)

fetchThreat

Agentic

Fetch detail for one specific identified threat. Get the threatId from listThreats.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • securityThreatModelId (required): The Security Threat Model ID, e.g. CJTY-ST-001. Get it from listSecurityThreatModel
  • threatId (required): The threatId (e.g. 'CJTY-TT-001')

Returns: Threat data as generic JSON (resourceId, resourceType, properties)

fetchStage

Agentic

Fetch one stage of a project's context assembly pipeline. Get the stageId from listStages.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • contextAssemblyId (required): The Context Assembly ID, e.g. CJTY-CA-001. Get it from listContextAssembly
  • stageId (required): The stageId (e.g. 'CJTY-SG-001')

Returns: Context Assembly Stage data as generic JSON (resourceId, resourceType, properties)

fetchRubric

Agentic

Fetch one specific evaluation rubric. Get the rubricId from listRubrics.

View Example

Parameters:

  • projectId (required): The project ID, e.g. CJTY
  • evalRubricId (required): The Eval Rubric ID, e.g. CJTY-ER-001. Get it from listEvalRubric
  • rubricId (required): The rubricId (e.g. 'CJTY-RB-001')

Returns: Rubric data as generic JSON (resourceId, resourceType, properties)

Common Workflows

🚀 Starting a New Feature

  1. Get Product Context
    Use fetchProductInfo to understand the overall product
  2. Review User Story
    Use fetchUserStory — the response tells you which companion tools to call next
  3. Dive Deeper
    Use fetchUserStoryAcceptanceCriteria, fetchUserStoryWireframes, or fetchTestCasesForUserStory as needed
  4. Check Tech Stack
    Use fetchProjectEnvironment for implementation details

📋 Discovery & Planning

  1. Verify Access
    Use getUserInfo to confirm your contracts
  2. Get Vision
    Use fetchVisionStatement for strategic direction
  3. List User Stories
    Use fetchUserStoriesForPersona to see all related work
  4. Check NDA Status
    Use checkNdaStatus if accessing confidential data

🎯 Quick Context Refresh

  1. Product Summary
    Use fetchElevatorPitch for quick overview
  2. Story Details
    Use fetchUserStory to get core requirements
  3. Technical Details
    Use fetchTestCasesForUserStory or fetchUserStoryWireframes for deeper context

Supported Development Tools

The Rezoomex MCP Server works with any MCP-compatible IDE

IDE / Tool MCP Support Transport Type Setup Difficulty
ChatGPT
Custom MCP server, in Developer mode 		✓ Native Streamable HTTP Easy
Claude
Custom connector, on the web and the desktop app 		✓ Native Streamable HTTP Easy
Cursor
AI-powered code editor 		✓ Native Streamable HTTP Easy
Windsurf
AI development environment 		✓ Native Streamable HTTP Easy
Claude Code
Official Claude CLI 		✓ Native Streamable HTTP Easy
OpenAI Codex
OpenAI's CLI tool 		✓ Native Streamable HTTP Easy
VS Code / GitHub Copilot
Copilot agent mode, via MCP 		✓ Extension Streamable HTTP Medium
Antigravity
Requires the mcp-remote proxy 		⚠ Via proxy HTTP via mcp-remote Medium
Zed
High-performance editor 		⚠ Beta Streamable HTTP Medium
Cline / Replit
Other MCP clients 		→ MCP Standard Streamable HTTP Medium

Note: Connect to https://mcp.rezoomex.com/mcp, which speaks Streamable HTTP. Any tool supporting the Model Context Protocol over HTTP can connect.

Development Environment Integrations

Connect your favorite tools to your contract-scoped workspace and project context.

ChatGPT

Your playbook in ChatGPT: personas, user stories, acceptance criteria and test cases

Claude

Your playbook in Claude, on the web and the desktop app, as a custom connector

VS Code

Full IDE integration with contract-scoped context and intelligent project navigation

Cursor

Enhanced AI coding experience with your specific contract requirements and acceptance criteria

Windsurf

Advanced AI-powered development environment with seamless contract-based workflow integration

OpenAI Codex

OpenAI's command-line coding assistant with native MCP support for contract-based development

Antigravity

Google's AI-powered IDE with MCP support for contract-based development workflows

Cline

Command-line access to playbook data, user stories, and contract specifications

Zed

Lightning-fast performance with integrated contract-based development workflows

Replit & Other IDEs

Connect any MCP-compatible development environment to your Rezoomex contracts

Integration Guide

Choose your development environment and connect to your contract-based workspace.

ChatGPT · custom MCP server

Add Rezoomex in ChatGPT as a custom MCP server. Our ChatGPT page has the values to enter, a link to OpenAI’s steps, and a prompt to check it works.

View the ChatGPT setup guide →
Claude · custom connector

Add Rezoomex in Claude as a custom connector; it works on the web and in the desktop app. Our Claude page has the values to enter, a link to Anthropic’s steps, and a prompt to check it works.

View the Claude setup guide →
cursor-mcp-config.json 
{
  "mcpServers": {
    "product-playbook": {
      "url": "https://mcp.rezoomex.com/mcp",
      "type": "http"
    }
  }
}

Important: Do not change the server name "product-playbook" in Cursor. This name is required for proper functionality.

windsurf-mcp-config.json 
{
  "mcpServers": {
    "product-playbook": {
      "url": "https://mcp.rezoomex.com/mcp",
      "type": "http"
    }
  }
}
terminal 
# Install the Rezoomex MCP Server
claude mcp add --transport http rezoomex-mcp https://mcp.rezoomex.com/mcp

# Verify installation
claude mcp list
terminal 
# Add and authenticate with the Rezoomex MCP Server
codex mcp add rezoomex-playbook --url https://mcp.rezoomex.com/mcp

# Verify connection
codex mcp list

Note: The codex mcp add command will automatically prompt you to authenticate via OAuth in your browser.

1. Open Settings JSON

Press Cmd+Shift+P (Mac) or Ctrl+Shift+P (Windows/Linux), then search for:

Command Palette 
Preferences: Open User Settings (JSON)

2. Add the MCP server config

Paste the following into your settings.json (merge with existing content — don't replace the whole file):

settings.json 
{
  "mcp": {
    "servers": {
      "rezoomex-playbook": {
        "type": "http",
        "url": "https://mcp.rezoomex.com/mcp"
      }
    }
  }
}

3. Authenticate

After saving, open the Command Palette again and run MCP: List Servers. Click the Rezoomex server and select Connect — VS Code will open your browser to sign in.

Antigravity has a known bug where OAuth tokens are not sent with remote MCP connections. Use mcp-remote as a local proxy — it handles the OAuth login flow and forwards requests to our server with your token automatically.

1. Add the server configuration

In Antigravity, open ... → MCP Servers → Manage MCP Servers → View raw config, or edit the file directly at ~/.gemini/antigravity/mcp_config.json:

~/.gemini/antigravity/mcp_config.json 
{
  "mcpServers": {
    "rezoomex-playbook": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://mcp.rezoomex.com/mcp"
      ]
    }
  }
}

2. Authenticate on first launch

Save the config and reload Antigravity. On first use, mcp-remote will open your browser to complete the login. After that, your token is stored in ~/.mcp-auth and refreshed automatically.

Requires Node.js. mcp-remote is downloaded automatically via npx — no manual install needed.

What's New

Recent feature releases — bug-fix-only versions are omitted

Current
v1.0.68
October 2026
  • New setup pages for ChatGPT and Claude: the exact values to enter, a link to the vendor’s own steps, a prompt to check the connection, and the plan limits worth knowing before you start.
  • Every setup guide now uses https://mcp.rezoomex.com/mcp. If you added Rezoomex with the older SSE address, it still works, but Claude is deprecating SSE — switch to /mcp when you next touch your config.
  • Claude Desktop no longer has its own setup tab: add Rezoomex once as a connector on the Claude page and it works on the web and in the desktop app. The setup guide now opens on Claude, and every tool shows its icon.
v1.0.67
October 2026
  • fetchPlaybookTree is now available to everyone, on every connection. It returns your whole project skeleton in one call — every saved resource id, its type and its parent, without the contents — so you can see what already exists before creating, updating or removing anything, instead of listing each parent in turn. It also surfaces resources whose parent has been deleted, which per-parent lookups cannot show. Project owners only.
  • getLastUsedIds is now available to everyone too. It reports the highest id number currently in use for each resource type, so you can mint the next one without guessing. Note it reports the highest id that still exists: if the highest-numbered resource was deleted, its number is reported as free. Project owners only.
v1.0.62
September 2026
  • listPersonas no longer reports a false "not found" for a project that simply has no personas saved yet — an empty project is a normal, non-error result, while a project that genuinely does not exist is still reported as an error
  • fetchPersona now returns a structuredContent object alongside the text (resourceId, parentId, resourceType, properties, isComplete), with properties passed through exactly as saved — empty strings, empty arrays, nulls and fields the text view does not render all survive, so you can read the exact saved state
  • listPersonas now returns one structured item per persona carrying its exact project-prefixed resourceId, plus name and role where saved — target a specific persona by ID instead of parsing text
  • Both tools now fail closed: a record that is malformed, belongs to another project, or is not a persona is reported as an error (MALFORMED_RESPONSE) instead of being returned as a partial or misleading result
  • Both connection types (SSE and Streamable HTTP) now advertise identical output schemas for these tools, so a client sees the same machine-readable shape whichever it uses

Frequently Asked Questions

Common questions about setup, authentication, and tool usage

How do I know if my IDE supports MCP?

Check if your IDE has native MCP support or an MCP extension. Look for "Model Context Protocol" or "MCP" in your IDE's documentation or extension marketplace.

The tools listed in our compatibility matrix above have confirmed MCP support with HTTP/SSE transport.

How do I authenticate with the MCP server?

After adding the MCP server configuration:

  1. Open your IDE's MCP settings/servers page
  2. Find the Rezoomex MCP server in the list
  3. Look for an "Authenticate" button or link (each IDE displays this differently)
  4. Click the authentication prompt - this will open your browser (if it doesn't open, copy the sign-in URL your IDE displays and open it manually)
  5. Sign in with your Rezoomex account credentials in the browser
  6. After successful login, close the browser tab and return to your IDE
  7. Verify the connection status shows as authenticated

IDE-Specific Notes:

  • ChatGPT: You sign in when you create the custom MCP server — see Connect Rezoomex to ChatGPT
  • Claude: Choose Sign in now when you add the connector — see Connect Rezoomex to Claude
  • Cursor: Check Settings → Features → MCP Servers
  • Windsurf: Check Settings → MCP Servers
  • Claude Code: Authentication happens automatically when you first use a tool
  • VS Code / GitHub Copilot: Open Command Palette → MCP: List Servers, click the server, then Connect
Where do I find my projectId?

Your projectId is provided in your contract onboarding email or can be found in the Rezoomex dashboard. Some tools also accept productCode as a legacy alias — it's the same value.

If you don't have your projectId, contact your project administrator or email team@rezoomex.com.

Why am I getting "Authentication failed" errors?

Common authentication issues:

  • Expired token: Restart your IDE to trigger a new authentication flow
  • Wrong credentials: Ensure you're using your Rezoomex account email and password
  • No active contracts: Verify you have active contracts in the Rezoomex dashboard
  • Firewall/proxy issues: Ensure your network allows connections to mcp.rezoomex.com and Auth0
Can I use multiple MCP servers in my IDE?

Yes! Most MCP-compatible IDEs support multiple MCP servers. Each server is configured with a unique name in your settings.

For example, you can have both Rezoomex and other MCP servers configured simultaneously.

What if my tool returns "Permission denied"?

Permission errors mean you don't have access to the requested resource:

  • Verify the projectId is for a contract you're assigned to
  • Use getUserInfo to check your access permissions
  • Use checkNdaStatus if accessing confidential data
  • Contact your project administrator to verify contract scope
What do I do when a tool returns an error?

Every error response includes a Suggested next steps section telling you exactly which tool to call to recover:

  • Not found errors: The response tells you which browsing tool to use — e.g. if a story ID is wrong, it suggests calling fetchUserStoriesForPersona to find valid IDs
  • NDA Required: Call checkNdaStatus to confirm your NDA status, then complete NDA signing in your Rezoomex workspace
  • Authentication failed: Re-authenticate via your MCP client settings — the session token may have expired
  • Network errors: Check your connection and retry the same tool call — no state is changed on failed requests
  • Acceptance criteria / wireframes / test cases missing: The tool will tell you to confirm the story exists first with fetchUserStory

Always read the full error response — it will tell you the exact next tool call to make.

How do I update my MCP server configuration?

To update your configuration:

  1. Locate your IDE's MCP configuration file (see setup guide above)
  2. Update the server URL or settings as needed
  3. Restart your IDE for changes to take effect

Note: Do not change the server name "product-playbook" in Cursor—this name is required.

Where can I get help if I'm stuck?

Support resources:

  • Email: team@rezoomex.com
  • Project Admin: Contact your contract project administrator
  • Documentation: Review the detailed guides in the setup section above
  • MCP Protocol: Visit modelcontextprotocol.io for protocol details

Found a bug or need help?

Your project admin can also help with access and contract permissions.

Email team@rezoomex.com