Projects & Campaigns

A project is a multi-phase campaign to discover primes in a specific form and range. Projects coordinate resources, track costs, and manage the lifecycle from initial scouting through record-breaking verification.

Project Lifecycle

┌──────────┐     ┌──────────┐     ┌──────────┐     ┌──────────┐
│  Scout   │────▶│  Search  │────▶│  Verify  │────▶│ Complete │
│          │     │          │     │          │     │          │
│ Quick    │     │ Full     │     │ Re-verify│     │ Archive  │
│ sweep to │     │ range    │     │ all PRPs,│     │ results, │
│ estimate │     │ search   │     │ generate │     │ submit   │
│ density  │     │ with AI  │     │ certs    │     │ records  │
└──────────┘     └──────────┘     └──────────┘     └──────────┘

Phase states

Phase	Purpose	AI Engine role
Scout	Quick sweep of a small range to estimate candidate density and cost	Calibrate cost model, set expectations
Search	Full-range search with allocated workers. Main compute phase	Dynamic resource allocation, stall detection
Verify	Re-verify all probable primes, generate deterministic certificates where possible	Prioritize high-value candidates for re-verification
Complete	Archive results, submit records to databases, write summary	Report generation, strategy learning

Configuration

Projects are defined via TOML configuration or the dashboard:

toml

[project]
name = "Factorial 200K+"
description = "Search for factorial primes above 200,000"
form = "factorial"

[target]
start = 200000
end = 300000
goal = "world_record"    # or "coverage", "count"

[budget]
max_compute_hours = 5000
max_cost_usd = 250.0
daily_limit_usd = 25.0

[resources]
min_workers = 4
max_workers = 16
priority = "high"        # high, normal, low

[phases]
scout_range = [200000, 201000]  # Quick density estimate
verify_all = true               # Re-verify all PRPs

Cost Estimation

Before committing resources, the project system estimates total cost using the AI engine's power-law cost model:

text

Project: Factorial 200K+
Form:    factorial
Range:   200,000 → 300,000

Estimated candidates:     ~8,500
Sieve survival rate:      ~12%
Candidates to test:       ~1,020
Avg test time:            45 min/candidate
Total compute:            ~765 core-hours
Estimated cost:           $38.25 (at $0.05/core-hour)
Expected primes:          0-2 (based on heuristic density)

World Record Tracking

Projects targeting world records automatically track the current standings by scraping t5k.org (The Prime Pages):

Current record holder for each form
Record digit count and discovery date
Gap analysis — how far the project's best result is from the record
Leaderboard position tracking over time

Orchestration

The project orchestrator runs a 30-second tick loop that manages phase transitions and resource allocation:

Auto-advance — Moves from Scout to Search when density estimate is confident
Auto-verify — Triggers Verify phase when Search reaches target coverage
Budget enforcement — Pauses searches if daily spend limit is reached
Worker scaling — Requests more workers from the AI engine when needed
Event logging — All phase transitions and decisions are recorded

Dashboard

Manage projects at app.darkreach.ai/projects:

Create and configure new projects
Monitor phase progress with timeline visualization
View cost breakdown and budget burn rate
See discovered primes and their verification status
Compare against world records in real time

API Endpoints

Method	Path	Description
GET	`/api/projects`	List all projects with status
POST	`/api/projects`	Create a new project
GET	`/api/projects/:id`	Get project details with phases and events
PUT	`/api/projects/:id`	Update project configuration
POST	`/api/projects/:id/advance`	Manually advance to next phase

Example: Creating a Campaign

bash

# Create a new factorial prime project via the API
curl -X POST https://api.darkreach.ai/api/projects \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Factorial 200K+",
    "form": "factorial",
    "config": {
      "target": { "start": 200000, "end": 300000 },
      "budget": { "max_cost_usd": 250.0 },
      "resources": { "min_workers": 4 }
    }
  }'