Agenten stärken
Die eigene URL des Herausgebers. Kein PlugMCP-Endpunkt.
Host a copy with our keys and logs.
https://career-workspace-ambujsystems.vercel.app/api/mcpDie Claude-Schaltfläche oben öffnet Claude mit ausgefülltem Namen und ausgefüllter URL. Sie bestätigen dort noch.
claude mcp add --transport http io.github.ambuj123-lab-covercraft-mcp https://career-workspace-ambujsystems.vercel.app/api/mcpRun in a terminal.
{ "mcpServers": { "io.github.ambuj123-lab-covercraft-mcp": { "command": "npx", "args": [ "-y", "mcp-remote", "https://career-workspace-ambujsystems.vercel.app/api/mcp" ] } }}claude_desktop_config.json
Remote-URL oder ein Befehl auf Ihrem Rechner? Worin sich die beiden Installationen unterscheiden
Auf PlugMCP ausführen
A new server you create. It is not this listing's official endpoint. Use the publisher install when you want their URL and their credentials.
Starten99% of cover letter tools on the internet are simple prompt wrappers that cause immediate disqualification during technical recruiter screens:
CoverCraft is not a template filler. It is an Evidence-Grounded Career Workspace & Claim-Validation Engine designed for senior engineers and discerning hiring teams.
Candidate Resume (PDF) Job Description (JD) Live Web (Tavily + Jina AI)
│ │ │
▼ ▼ ▼
┌──────────────────┐ ┌──────────────────┐ ┌────────────────────┐
│ Deterministic │ │ Deterministic │ │ 8-Page Multi-Board │
│ Text & Evidence │ │ Fallback Regex │ │ Crawl & Markdown │
│ Extraction Layer │ │ & Vendor Auditor │ │ Deep Extraction │
└─────────┬────────┘ └────────┬─────────┘ └─────────┬──────────┘
│ │ │
└─────────────────────────────┼─────────────────────────────────┘
│
▼
┌─────────────────────────────────┐
│ Human-in-the-Loop Gateway │
│ (Source Approval & Selection) │
└───────────────┬─────────────────┘
│
▼
┌─────────────────────────────────┐
│ Evidence-Grounded Generator │
│ • Strict Verbatim Anchors │
│ • Evidence-Grounded Claim Validation │
│ • Citation Footnotes [1], [2] │
└───────────────┬─────────────────┘
│
▼
┌─────────────────────────────────┐
│ Interview-Defensible Output │
│ (Every claim backed by facts) │
└─────────────────────────────────┘
careers, linkedin, naukri, and ambitionbox.r.jina.ai) Deep Scraping: Strips HTML clutter, popups, and cookie walls, returning clean, high-fidelity markdown with an exponential timeout fallback.When LLM JSON structures are delayed or partially populated, a deterministic regex pipeline extracts ground truth directly from raw JD tokens:
"Not verified from available sources".[1], [2].[Resume Anchor] & [Source Citation]) allow candidates to hover and verify exact source quotes with a live [Evidence Markers: ON/OFF] toggle. When candidates click Copy to Clipboard or Download Markdown/PDF, internal citation tags are automatically stripped to deliver 100% clean, submission-ready letters for recruiters.01. Competency Match → 02. MCP Company Research → 03. Human Approval → 04. Evidence Synthesis) with a live counting timer (00:04s) and active tool indicator so candidates clearly follow live execution without confusion or perceived stalls.[!TIP] 🛡️ Triple-Registry Verified FastMCP Server (Glama · Smithery · M8ven):
CoverCraft's FastMCP server and its 8 registered tools have been independently audited and verified across leading registries with Glama Grade A - A, M8ven Trust Score 89/100 (Grade B), and Smithery Verified (8 Tools) with continuous live monitoring and 0 security vulnerabilities.
mcp-server/).company_research: Real-time web search via Tavily with traceable source citations.evidence_validator: Claim Ledger validation against source data returning VERIFIED / PARTIAL / UNSUPPORTED.jd_analyzer: Evidence-backed JD matching against resume with actual proof anchors.ats_readiness: Deterministic heuristic audit on keyword coverage, format integrity, and length.source_filter: Algorithmic domain authority tiering, noise suppression, and Outdated News & Stale Tech Filter (<9m company initiatives, <18m engineering stack, [Historical Context] tag).cover_letter_generator: Evidence-grounded synthesis with structured in-line citation markers and zero-overclaim enforcement.github_proofer: Real-time candidate GitHub portfolio & commit history verification engine ($0 free tier PAT, 5,000 req/hr). Inspects public repositories, extracts commit SHAs and messages, and anchors resume technical claims directly into verifiable code evidence with zero overclaiming.huggingface_proofer: Autonomous GenAI model & weights verification engine ($0 free tier HF Token, 30,000 req/hr). Inspects candidate's public Hugging Face models, spaces, datasets, and aggregate download metrics (e.g. 700+ model downloads, qLoRA fine-tuning, GGUF quantization), cross-referencing AI claims against verifiable weights.stdio_server or deploy as an independent streaming microservice using --transport=sse.mcp-server/tools/github_proofer.py) and Next.js route /api/github-verify.sha: 7f2a1b9), commit messages, repository descriptions, and architecture files to ground technical claims in verifiable code history.mcp-server/tools/huggingface_proofer.py) and Next.js route /api/hf-verify.ambuj-ai-chatbot, legal-india-chatbot), mathematically verifying live model deployment and edge inference capabilities./api/parse-resume): Auto-extracts candidate Name, Email, LinkedIn URL, and GitHub Profile URL from uploaded PDF resumes via regex + LLM extraction, pre-filling the generator form.InputForm.jsx): Automatically detects target Role and Company Name as soon as a user pastes a Job Description, eliminating redundant manual typing.nextjs:nodejs user for zero root vulnerabilities.sequenceDiagram
autonumber
actor Candidate as User / Recruiter
participant Web as Next.js 16 Frontend
participant API as Route Handlers (/api/*)
participant Tavily as Tavily Search Engine
participant Jina as Jina AI Reader
participant GitHub as GitHub MCP Proofer
participant Langfuse as Langfuse Observability
participant LLM as Google Gemini 2.5 Flash Lite
participant DB as MongoDB Atlas
Candidate->>Web: Upload Resume (PDF) & Paste Job Description
Web->>API: POST /api/parse-resume & /api/analyze
API->>API: Auto-Extract Candidate Name, Email, LinkedIn & GitHub Handle
API->>API: Auto-Detect Target Role & Company from JD
API->>GitHub: GET /api/github-verify (MCP Tool #7 Commit Auditor)
GitHub-->>API: Verified Public Repositories, Commit SHAs & Timestamps
API->>LLM: Schema-Enforced Extraction (Skills & Match Matrix)
LLM-->>API: Structured Match Matrix & Vendor Indicators
API->>Langfuse: Log Trace & Child Spans (Latency, Tokens, Model)
API-->>Web: Render Evidence Match, GitHub Code Proof, Hugging Face Proof & 4-Tier Job Context Tabs
Candidate->>Web: Trigger Real-Time Research
Web->>API: POST /api/research { company, role }
API->>Tavily: Crawl 8 Authoritative URLs (Careers, News, Reviews)
Tavily-->>API: Raw Search Results
API->>Jina: Deep Scrape Top URLs to Clean Markdown
Jina-->>API: High-Fidelity Content
API-->>Web: Display HITL Source Cards (User Approves Sources)
Candidate->>Web: Select Tone & Click "Generate Cover Letter"
Web->>API: POST /api/generate { approvedSources, evidenceMap }
API->>LLM: Evidence-Grounded Generation Prompt (Claim Validation)
LLM-->>API: Streamed Verified Cover Letter + Citations
API->>DB: Log Telemetry & Audit Record
API-->>Web: Render Final Letter with Copy, Download & Interview Defense
career-workspace-ambujsystems/
├── web/ # Next.js 16 Fullstack Application
│ ├── public/ # Static assets & SVG icons
│ ├── src/
│ │ ├── app/
│ │ │ ├── api/
│ │ │ │ ├── analyze/ # Resume parsing & JD matching handler
│ │ │ │ ├── auth/ # NextAuth Google OAuth endpoints
│ │ │ │ ├── generate/ # Evidence-grounded generation pipeline
│ │ │ │ └── research/ # Tavily 8-page crawl + Jina Reader pipeline
│ │ │ ├── docs/ # Interactive Architectural Documentation
│ │ │ ├── generate/ # Main interactive workspace interface
│ │ │ ├── globals.css # Design system tokens, tactile spring curves & glow utilities
│ │ │ ├── layout.js # Root layout with NextAuth Provider
│ │ │ └── page.js # High-conversion Obsidian Black landing page
│ │ ├── components/
│ │ │ ├── ResultsPanel.jsx # 4-tier tabbed intel & vendor verification UI
│ │ │ ├── SourceCard.jsx # Human-in-the-loop approved source cards
│ │ │ └── ... # Reusable UI component modules
│ │ ├── lib/
│ │ │ ├── gemini.js # Resilient Google Gemini client with retry
│ │ │ └── mongodb.js # Production connection-cached MongoDB client
│ │ └── prompts/ # Strict system prompts & JSON schemas
│ ├── Dockerfile # Multi-stage standalone Next.js container (~120MB)
│ ├── next.config.mjs # Standalone output configuration
│ └── package.json # Dependencies & build scripts
├── mcp-server/ # Anthropic Model Context Protocol (MCP)
│ ├── server.py # Stdio protocol handler with analysis tools
│ ├── Dockerfile # Python 3.11-slim container definition
│ └── requirements.txt # MCP protocol SDK & utilities
├── docker-compose.yml # Multi-service local orchestrator
└── README.md # Production Engineering Specification
v20.x or v22.xgit clone https://github.com/Ambuj123-lab/career-workspace-ambujsystems.git
cd career-workspace-ambujsystems/web
Create .env.local inside the web/ folder (or copy .env.example from repository root):
# --- Core AI & Grounding Search (Required) ---
GEMINI_API_KEY=your_gemini_api_key_here
TAVILY_API_KEY=your_tavily_api_key_here
# --- Authentication & Security (NextAuth.js) ---
GOOGLE_CLIENT_ID=your_google_client_id_here
GOOGLE_CLIENT_SECRET=your_google_client_secret_here
NEXTAUTH_URL=http://localhost:3000
NEXTAUTH_SECRET=your_random_32_character_secret_here
# --- Candidate Proofing & Verification (Optional) ---
GITHUB_TOKEN=your_github_token_here
HF_TOKEN=your_huggingface_token_here
JINA_API_KEY=your_jina_api_key_here
# --- LLM Provider Fallback & Telemetry (Optional) ---
OPENROUTER_API_KEY=your_openrouter_api_key_here
MONGODB_URI=mongodb+srv://<username>:<password>@cluster.mongodb.net/?appName=CoverCraft
MONGODB_DB_NAME=covercraft_db
npm install
npm run dev
Open http://localhost:3000 in your browser.
Build and run the multi-stage standalone container locally:
# Build multi-stage image (~120MB footprint)
docker build -t covercraft-web:latest ./web
# Run standalone container
docker run -p 3000:3000 --env-file ./web/.env.local covercraft-web:latest
Or deploy both the Next.js app and the Python MCP server via Docker Compose:
docker compose up -d --build
Generative AI Engineer & Production Systems Specialist
Building production-grade agentic architectures, deterministic extraction pipelines, and high-reliability LLM systems.
This project is open-source and licensed under the MIT License.
Architected with ❤️ by Ambuj Kumar Tripathi • Powered by Google Gemini 3.5 & 3-Tier Multi-Provider Cascade (OpenRouter Nemotron 550B) • Langfuse Telemetry • Tavily AI • Jina AI • Next.js 16
Live Hugging Face model proofing, ATS resume scoring and career tools by Ambuj Kumar Tripathi
{ "mcpServers": { "io.github.ambuj123-lab-covercraft-mcp": { "url": "https://career-workspace-ambujsystems.vercel.app/api/mcp", "headers": {} } }}Tools
Noch keine Tool-Liste. Remote-Server erhalten eine nach dem ersten Health-Check.
FAQ
Kopieren Sie auf dieser Seite die http-Konfiguration für Cursor, Claude Desktop, Claude Code oder VS Code. Sie nutzt die Installation des Herausgebers, keinen PlugMCP-Proxy.
In diesem Eintrag ist keine erforderliche Umgebungsvariable aufgeführt.
CoverCraft AI By Ambuj Kumar Tripathi wird über HTTP erreicht.
Zuletzt geprüft: 2026-11-08. Live Hugging Face model proofing, ATS resume scoring and career tools by Ambuj Kumar Tripathi