Enhancv launches an MCP connector to prepare for jobs
Enhancv launched a hosted remote MCP server over Streamable HTTP that enables AI assistants to edit resumes live with OAuth 2.1 and PKCE authentication.
By Dillip Chowdary • Nov 16, 2026 • Source: github.com

At a glance
- Product
- Enhancv MCP Server
- Registry ID
- com.enhancv/mcp
- Endpoint
- https://api.enhancv.com/mcp
- Auth standard
- OAuth 2.1 with PKCE
- Pricing tiers
- Free read-only; paid editing and creation
- Rate limits
- 60 calls per minute and 1,000 per hour per user
In this post
Enhancv launched an official remote Model Context Protocol server connecting conversational assistants and coding agents directly to user resumes, cover letters, and job trackers. According to github.com's report, the integration allows external artificial intelligence clients to inspect, modify, and restyle career documents without requiring local daemon installations.
This report examines the remote endpoint architecture, authentication standards, client configuration commands, granular tools, and security boundaries. It explains how job seekers and engineers can connect AI agents to cloud canvases while keeping account credentials and billing records isolated.
01Remote server architecture and streamable HTTP transport
The server operates on Enhancv cloud infrastructure via a remote endpoint at https://api.enhancv.com/mcp. Users do not need to install local packages, run background services, or manage private server hardware. External client applications communicate directly with the cloud service over standard web connections.
Connection transport uses Streamable HTTP alongside OAuth 2.1 authentication with Proof Key for Code Exchange protocol standards. The system supports dynamic client registration and client identity metadata documents for secure sessions. The connector is indexed as com.enhancv/mcp within the official MCP Registry.
Server documentation and client configuration schemas reside in a public repository released under the MIT License. Enhancv provides technical support for connector setup through its help@enhancv.com inbox. This architecture guarantees that third-party applications integrate cleanly while document rendering remains centralized.
02Live editor synchronization and document tailoring
The connector enables AI models to evaluate resumes against vacancy descriptions to tailor candidate profiles for specific positions. Assistants can rewrite summaries, polish bullet points, and update technical skills while keeping every detail true to factual experience. Users can also draft cover letters from resume data and link them to saved job postings.
When an applicant keeps a document open in the web application, edits made by connected assistants appear live on the canvas. This immediate visual update lets users monitor adjustments as they occur and decide whether to accept or undo each modification. The interface maintains human review across all automated drafting tasks.
Styling tools allow models to modify document presentation without corrupting underlying section layouts. Connected agents can adjust typography fonts, alter margins, switch layout templates, and reorder structural sections. Completed versions can be exported directly from chat environments as downloadable PDF files.
03Assistant connection workflows and configuration commands
Advertisement
Tech Pulse Daily
Get tomorrow's pulse first
Join engineers who read Tech Pulse before stand-up. Free, weekday mornings.
Chat interfaces link through standard connector settings using the remote endpoint URL https://api.enhancv.com/mcp. Users operating Claude on web, desktop, or mobile platforms sign in to Enhancv to authorize the integration. ChatGPT users complete an identical connection process within plugin settings to establish account communication.
Developer environments support setup through terminal commands and configuration files. Claude Code registers the service using claude mcp add with user scope and HTTP transport flags. Visual Studio Code links via code --add-mcp, while Cursor users define the endpoint inside their local mcp.json file.
Generic clients supporting Streamable HTTP register the server endpoint and launch authentication prompts upon first use. The web interface also provides an Account AI connections page featuring one-click setup links for supported tools. Users retain full control to revoke connections at any time.
04Granular tool catalog and optimistic versioning
The server provides inspection tools guarded by the resumes:read authorization scope. Available read methods include list_resumes to sort documents by edit date, get_resume to fetch field identifiers, and download_resume to render PDF links. Read actions also feature list_cover_letters, list_saved_jobs, get_saved_job, and search_jobs for feed discovery.
Write operations operate under the separate resumes:write permission scope. The write catalog includes edit_field, edit_multiple_fields, edit_list, add_item, remove_item, add_section, delete_section, move_section, and move_item. Document creation and tracking tools include create_resume, duplicate_resume, import_resume, create_cover_letter, save_job, and update_saved_job.
Every edit operation applies optimistic concurrency tracking to avoid overwriting simultaneous changes. Mutations require the document's current version identifier before modifying fields. If the resume changed in the interim, the server rejects the edit and returns the latest version so the model can inspect it and retry.
05Subscription access tiers and rate limits
Feature access differs between free accounts and active paid subscriptions across the platform. Free accounts allow assistants to read resumes, view cover letters, inspect saved jobs, and search vacancies. Paid subscriptions unlock all write tools, including content generation, section creation, document duplication, and PDF rendering.
| Plan tier | Permitted capabilities | Tool permission scopes |
|---|---|---|
| Free Enhancv plan | Read resumes, inspect cover letters, review saved jobs, search feed | Read-only operations under resumes:read |
| Paid Enhancv plans | Edit fields, create documents, import text, export PDF, update tracker | Full read and write operations under resumes:write |
Strict security boundaries isolate connected assistants from sensitive account settings and billing options. The integration never exposes user passwords and prohibits models from deleting resumes or removing saved jobs. Users can disconnect any application immediately within the web dashboard under Account AI connections.
Note: Operational limits restrict usage to 60 calls per minute and 1,000 calls per hour per user across all connected tools. Account holders can immediately sever access for any external application by navigating to Account and choosing AI connections inside the Enhancv portal.
Frequently asked questions
What endpoint connects AI assistants to the Enhancv MCP server?
Assistants connect to https://api.enhancv.com/mcp using Streamable HTTP. Users authenticate using OAuth 2.1 with PKCE by signing in to their Enhancv account.
Can free Enhancv accounts edit resumes using AI tools?
Free plans only support read operations to inspect resumes, cover letters, and saved jobs. Editing fields, creating new documents, and downloading PDFs require a paid subscription.
How does the server prevent data loss during concurrent edits?
Tools require the document's current version identifier before applying revisions. If a resume changed beforehand, the tool returns the latest version so the AI can evaluate changes and retry.
What rate limits apply to Enhancv MCP tool calls?
Tool calls are capped at 60 requests per minute and 1,000 requests per hour per user. Users can disconnect integrations at any time under Account AI connections.
Sources
Author
Dillip Chowdary
Writes Tech Bytes coverage of AI, engineering, and the tools that actually ship. Editor of Tech Pulse Daily.
Related on Tech Bytes
Advertisement