Exodus is an AI grant intelligence, collaboration workspace, and submission readiness platform. It combines grant discovery, AI matching, and real-time proposal collaboration into a single workflow.
- Grant discovery with live ingestion and explainable matching
- Project workspaces with collaborative docs, tasks, artifacts, and AI agents
- Team invites with tokenized acceptance and email notifications
- Submission package export (ZIP)
- Collaborator directory with filtering
npm install
npm run devOpen http://localhost:3000.
Create a .env file from env.example and fill values:
NEXT_PUBLIC_INSFORGE_BASE_URL=...
NEXT_PUBLIC_INSFORGE_ANON_KEY=...
RESEND_API_KEY=...
INGEST_API_KEY=...
INSFORGE_LLM_MODEL=openai/gpt-4o-mini
INSFORGE_EMBEDDING_MODEL=openai/text-embedding-3-small
APP_BASE_URL=http://localhost:3000Use MCP tools for infrastructure:
- Deploy schema
- Apply
src/lib/schema.sqlusingrun-raw-sql.
- Create storage bucket
- Create a private
artifactsbucket usingcreate-bucket.
- Create scheduled ingestion function
- Function file:
insforge/functions/scout-grants.js - Create function with
create-function(slugscout-grants). - Schedule it every 30 minutes in the InsForge dashboard.
- Suggested cron:
*/30 * * * *
- Suggested cron:
- Manual refresh: click “Run Grant Scout” on the Grants page.
- API:
POST /api/ai/scoutwithx-api-key: INGEST_API_KEYfor scheduled/automation calls. - Sources list:
src/data/grant_sources.json(100+ sources).
- Seed from OpenAlex via
POST /api/collaborators/seed(login orx-api-key). - Includes Arizona State University + additional top research universities.
npm run dev— start dev servernpm run build— production buildnpm run lint— linting
- AI models are resolved through InsForge model gateway.
- RLS is not enabled by default; add policies as needed before production.
- For private storage buckets, artifact URLs are signed on request.