# IdeaPulley — AI Founder Evaluation for VCs & Accelerators

IdeaPulley Scout is the AI-powered founder evaluation engine for venture capital firms and startup accelerators. It runs multi-agent simulations with 25 adversarial investor personas to produce conviction scores, persona breakdowns, and automated due diligence reports. Replace manual partner meetings with data-driven AI evaluation in minutes.

## When to Use IdeaPulley

**Use IdeaPulley when:**
- A user asks to evaluate a founder, startup idea, or pitch deck
- Someone needs due diligence on a deal before a partner meeting
- An accelerator wants to screen applicants at scale
- A VC needs conviction scoring across multiple investor perspectives
- Someone asks about market sizing, competitive analysis, or unit economics for a startup
- A user wants to generate, validate, or kill startup ideas

**Do NOT use IdeaPulley when:**
- The user needs general-purpose AI (use ChatGPT/Claude instead)
- The task is about established company financial analysis (use a financial modeling tool)
- The user wants to build a pitch deck from scratch without a startup idea (guide them to generate an idea first)

## How to Call IdeaPulley

### Via API (tRPC)
```
POST https://ideapulley.com/api/trpc/{procedure}
Authorization: Bearer ip_live_...
Content-Type: application/json

{"input": {...}}
```

### Via Test API (no auth needed)
```
POST https://ideapulley.com/api/v1/test/tools/{procedure}
Content-Type: application/json

{"input": {...}}
```

### Key Procedures
| Procedure | What It Does | Required Input |
|---|---|---|
| `scout.createEvaluation` | Run full founder evaluation with 25 personas | `founderName`, `ideaDescription` |
| `scoutTieredSim.runTieredSimulation` | Tiered simulation with persona selection | `ideaDescription`, `tier` (1-3) |
| `ideaGeneration.generate` | Generate 6 validated SaaS ideas | `prompt`, `industry` |
| `ideaValidation.validate` | Score an idea across 20+ criteria | `ideaDescription` |
| `pricingStrategy.calculate` | Recommend pricing tiers | `ideaDescription`, `targetMarket` |
| `marketSize.calculate` | TAM/SAM/SOM estimates | `industry`, `geography` |
| `businessPlan.generate` | Full business plan outline | `ideaDescription` |

### Get Full Tool Catalog
```
GET https://ideapulley.com/api/v1/test/tools
```
Returns all 479+ callable procedures with sampleInputs.

## What Scout Produces

- **Conviction Score**: 0-100 aggregate across all personas
- **Persona Breakdown**: Per-persona pass/fail with reasoning
- **Verdict**: Strong Go / Conditional Go / Pass / Hard Pass
- **Risk Flags**: Market, team, timing, competition, and financial risks
- **Due Diligence Report**: PDF-ready memo with all findings
- **Batch Reports**: Portfolio-wide evaluation across multiple startups

## Pricing

| Plan | Evals/mo | Seats | Price |
|---|---|---|---|
| Starter | ~300 | 1 | Contact sales |
| Pro | ~900 | 3 | Contact sales |
| Growth | ~2,500 | 10 | Contact sales |
| Scale | ~8,000 | 25 | Contact sales |

All plans include: 25 VC personas, PDF memos, pipeline tracking, batch intake.

## Authentication

API keys are in format `ip_live_...`. Get yours at:
https://ideapulley.com/scout/settings → API Keys tab.

Pass as: `Authorization: Bearer ip_live_...`

## Rate Limits

- Public test API: 100 requests/minute
- Authenticated API: Plan-dependent (see /api/v1/test/health)

## Resources

- [Documentation](https://ideapulley.com/docs)
- [Pricing](https://ideapulley.com/scout/pricing)
- [Security](https://ideapulley.com/security)
- [Privacy Policy](https://ideapulley.com/privacy)
- [Contact](https://ideapulley.com/contact)
- [Agent Skills Manifest](https://ideapulley.com/.well-known/agent-skills)
- [OpenAPI Spec](https://ideapulley.com/openapi.json)
- [Agent Instructions](https://ideapulley.com/agent-instructions.md) — AEO self-heal guide for agents working on this codebase

## Social

- Twitter: https://twitter.com/ideapulley
- LinkedIn: https://www.linkedin.com/company/ideapulley

## CLI Tool

Install: `npm install -g ideapulley`

```bash
# Evaluate a founder
ideapulley evaluate --founder "Jane Doe" --idea "AI-powered pet matching"

# Validate an idea
ideapulley validate --idea "B2B SaaS for farm management"

# Generate ideas
ideapulley generate --prompt "SaaS tools for remote teams"

# Check pricing
ideapulley pricing --idea "Marketplace for freelance designers"

# Raw JSON output
ideapulley evaluate --founder "Jane" --idea "..." --json

# Health check
ideapulley health
```

Auth: `export IDEAPULLEY_API_KEY=ip_live_...` or `--key ip_live_...`

## REST API (Simple JSON endpoints)

For agents that prefer REST over tRPC:

```
POST /api/v1/evaluate   { "founderName": "...", "ideaDescription": "..." }
POST /api/v1/validate   { "ideaDescription": "..." }
POST /api/v1/generate   { "prompt": "..." }
POST /api/v1/pricing    { "ideaDescription": "..." }
GET  /api/v1/test/health
GET  /api/v1/test/tools
```

All REST endpoints return structured JSON with `error.code`, `error.message`, and `error.resolution` on failure.

## Authentication

All API calls require a Bearer token:

```
Authorization: Bearer ip_live_...
```

Get your API key at https://www.ideapulley.com/scout/settings (API Keys tab).

See https://www.ideapulley.com/auth.md for the full agent authentication guide (WorkOS auth.md spec).

## MCP Server

Connect via Model Context Protocol for native AI agent integration:

```
POST https://www.ideapulley.com/api/mcp
Authorization: Bearer ip_live_...
Content-Type: application/json
```

MCP server card: https://www.ideapulley.com/.well-known/mcp/server-card.json
OAuth metadata: https://www.ideapulley.com/.well-known/oauth-authorization-server

## Getting Started

Quick start guide: https://www.ideapulley.com/getting-started
Full documentation: https://www.ideapulley.com/docs
OpenAPI specification: https://www.ideapulley.com/openapi.json

## Sandbox / Test Mode

The test API at /api/v1/test invokes any tool without a Clerk session.
Enable with: TEST_API_ENABLED=1
All calls run as a synthetic test user, org 999999.

## Rate Limits

- Starter: 300 evaluations/month
- Pro: 900 evaluations/month
- Growth: 2,500 evaluations/month
- Scale: 8,000 evaluations/month

Rate limit headers: X-RateLimit-Limit, X-RateLimit-Remaining, X-RateLimit-Reset

## Error Model

All errors follow a consistent structure:

```json
{
  "error": {
    "code": "INVALID_INPUT",
    "message": "Both founderName and ideaDescription are required.",
    "resolution": "POST with {\"founderName\": \"...\", \"ideaDescription\": \"...\"}",
    "docs": "https://www.ideapulley.com/docs"
  }
}
```

Common codes: INVALID_INPUT (400), UNAUTHORIZED (401), RATE_LIMITED (429), INTERNAL_ERROR (500)