grad-agent
An autonomous MCP agent that helps you apply to fully funded MS and PhD programs by discovering professors, verifying faculty status, matching projects, and drafting cold emails.
README
grad-agent
An autonomous MCP agent that helps you apply to fully funded MS and PhD programs.
- Discovers professors on arXiv in your research areas
- Verifies each candidate is actually faculty (Semantic Scholar, h-index, papers)
- Scrapes their lab page for a recruiting signal + email address
- Matches them to your strongest shipped project
- Drafts a specific, fact-checked cold email (Claude Haiku authors + verifies claims against paper abstracts)
- Compiles a per-school SOP to PDF (LaTeX)
- Tracks everything in xlsx: outreach, LOR requests, program deadlines
- Emails every draft to your inbox for review; nothing is sent to a professor without you
Runs as a stdio MCP server for Claude Code / Claude Desktop, or as a plain CLI.
Install
pipx install grad-agent
# or, if you prefer pip in a venv
python3 -m venv .venv && source .venv/bin/activate
pip install grad-agent
Set up in 5 minutes
grad-agent init
This writes:
~/.grad-agent/profile.yaml— your identity, projects, preferences~/.grad-agent/programs.yaml— target programs (seeded)~/.grad-agent/.env— secrets template
Fill in ~/.grad-agent/.env:
ANTHROPIC_API_KEY=sk-ant-...
SMTP_SERVER=smtp.gmail.com
SMTP_PORT=587
SMTP_USERNAME=you@gmail.com
SMTP_PASSWORD=<gmail app password>
SMTP_FROM=you@gmail.com
# Optional:
GITHUB_USERNAME=your-gh
GITHUB_TOKEN=github_pat_...
HF_USERNAME=your-hf
Fill in the important bits of ~/.grad-agent/profile.yaml:
name,identity_line,portfoliocv_path,transcript_path(absolute paths)degree_status: bachelors | masters(drives PhD eligibility gating)target_term,target_degreeresearch_areas: [nlp, ai4health, ...]seed_projects:3 to 10 flagship projects withname,pitch,link,tags
Then:
grad-agent sync # scan projects (GitHub + HF + local)
grad-agent run # one batch, drafts land in your inbox
Register with Claude Code
grad-agent register-claude
Prints the exact claude mcp add command. Once registered, /mcp in Claude Code shows grad-agent with ~25 tools.
Register with Claude Desktop
Edit ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"grad-agent": {
"command": "grad-agent",
"args": ["server"]
}
}
}
Daily autonomous run (macOS)
The package ships a launchd plist template. To fire every day at 08:00:
cp /path/to/grad_agent/templates/com.gradagent.daily.plist ~/Library/LaunchAgents/
launchctl load ~/Library/LaunchAgents/com.gradagent.daily.plist
Every morning: 3 verified faculty leads, hooks fact-checked against paper abstracts, in your inbox for review.
What each MCP tool does
| Tool | Purpose |
|---|---|
run_daily_batch(n, area) |
Full pipeline: verify → recruiting → hook + verify → draft → log |
outreach_log_view(limit) |
Show last N rows of the outreach xlsx |
outreach_mark_sent(prof, uni) |
Flag a row as actually sent to the prof |
sync_catalog(source) |
Pull projects from github, hf, or local |
list_projects_in_catalog() |
Show every project the matcher can see |
list_programs() |
Your target programs |
upcoming_deadlines(days) |
Any program deadline in the next N days |
lor_add / lor_outstanding / lor_mark |
Recommendation-letter tracker |
draft_cold_email(...) |
Manual per-prof draft |
draft_sop(...) |
Compile a Columbia-style SOP PDF for one school |
send_draft_to_me(path) |
Ship any draft file to your review inbox |
discover_profs(area) |
arXiv + OpenReview scan (raw candidates, no verification) |
Blog publishing tools (publish_article, update_article, ...) are gated behind blog.enabled: true in profile.yaml and are specific to the author's Turso-backed Next.js portfolio. Most users can ignore them.
What the agent will not do
- Send any email to a professor. Every send is manual, from your Gmail, after you read the draft.
- Fabricate a paper claim. The hook goes through a second Claude call that rejects any claim not present in the abstracts, and rewrites.
- Draft for programs you are ineligible for. If your
degree_statusisbachelors, PhD programs that require an MSc first are filtered out.
Requirements
- Python 3.10+
- macOS or Linux (Windows untested)
pdflatexon PATH if you want SOP PDFs (macOS: MacTeX; Ubuntu:texlive-latex-recommended)- Anthropic API key
- Gmail (or another SMTP) for the review-mailer
Where your data lives
Everything is under ~/.grad-agent/ by default, or $GRAD_AGENT_HOME if set:
~/.grad-agent/
profile.yaml identity + preferences
programs.yaml target programs
.env secrets (gitignored)
data/
outreach_log.xlsx every prof surfaced or drafted
lor_log.xlsx recommendation-letter tracker
catalog.json synced projects (GitHub + HF + local)
db.sqlite drafts + status
drafts/ per-school SOP + email drafts
Contributing
MIT licensed. PRs welcome for: more program templates, non-Gmail SMTP presets, non-arXiv source adapters, and better prof-verification heuristics.
Author
Kwabena Obeng · i-ninte.github.io/portfolio/
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.