Prompt Operations Platform

Prompts are living programs.

Author with research. Evaluate with a council of models. Evolve against real data. Ship to Claude, ChatGPT, or any API in one click.

Not sure where to start? Build one in 90 seconds →

Why not just another library

Static prompt libraries lose to the models themselves now.

We don't compete on template count. We compete on measurable quality — prompts that improve under evaluation and ship with a score.

Every other library

Static text. No feedback.

prompts/seo-brief.md
You are an SEO expert. Write a...
— last updated 2 years ago
— no evaluation data
— no improvement loop
Live
Promptseng

Scored. Evolving. Measurable.

agreement_score0.71 → 0.88 ▲
variant_v4proposed
variant_v3promoted · +14%
runs_last_7d2,418
# gets smarter while you sleep

How it works

Author · Evaluate · Evolve · Ship

01
Start with a research-backed template

Every seed prompt is tagged with the techniques it uses.

CoTSelf-VerifyHAPO
02
Fill variables; run Council

N models in parallel. Synthesis + dissent map, not a winner bracket.

GPT-4oClaude 3.5Gemini 2.0
03
Watch it evolve

Nightly variant generation. Judged against real examples. Public changelog.

variantsjudgedpromoted
04
Ship anywhere

Claude Skill, Custom GPT, MCP server, or a plain HTTPS API.

SkillGPTMCPAPI

60 prompts. All measured. All improving.

Every prompt carries its agreement score, its Council run history, and its Evolution lineage. No template graveyards.

7 PROMPTS

Marketing & Growth

avg effectiveness: 0.82
6 PROMPTS

Software Development

avg effectiveness: 0.82
5 PROMPTS

Sales & Business Dev

avg effectiveness: 0.82
4 PROMPTS

Finance & Banking

avg effectiveness: 0.82
4 PROMPTS

Human Resources

avg effectiveness: 0.82
4 PROMPTS

Operations & Logistics

avg effectiveness: 0.82

Enterprise-Grade Security

Built for teams who take security seriously. Your prompts, your keys, your control.

BYOK, always

Your API keys stay yours. We use AES-256-GCM with keys derived via scrypt. We never see your model outputs.

Encrypted at rest

All prompts, keys, and Council run history are encrypted in Postgres with row-level security.

No model proxy

We don't proxy your traffic. Your calls go directly to OpenAI, Anthropic, or Google from our edge.

GDPR ready

Export or delete your data anytime. Audit logs available on Team and Enterprise.

Pricing

Free is permanent. BYOK forever.

No trials. No credit card for Free. You pay when your team needs shared context, analytics, or SSO.

Free
$0forever
  • Browse the full library
  • 20 Council runs / month, 1 model
  • BYOK — your keys, your calls
  • Public Evolution changelog
Start Free
Recommended
Pro
$20/mo
  • Unlimited Council runs (BYOK)
  • Forge linter + scaffold injection
  • Evolution subscriptions
  • Export to Claude Skill / GPT / MCP
Team
$50/user/mo
  • Shared Context Capsules
  • Team prompt library
  • Semantic + hybrid search
  • Usage analytics
Enterprise
Custom
  • SSO / SCIM
  • On-prem or VPC deploy
  • Audit logs + SLAs
  • Custom evals & red-teaming
Talk to us

Prompts that get smarter
while you sleep.

Start Free (BYOK)

No trial. No credit card. Your keys, your calls.