System Overview // Foundry v1

Micro-Agent
Foundry

Autonomous agent chains that build, test, and improve themselves. No hardcoded vendors. No human in the loop. Just typed contracts, tournament selection, and relentless self-optimization.

Text Generators Are Not Agents

Agents that just generate text are useless. Real agents need to ACT: scrape, send, track, evaluate. Each step needs the best possible approach -- not hardcoded vendors.

Hardcoded Vendors Lock You In

Today's best email API might be tomorrow's worst. Agents wired to a single provider can't adapt when the landscape shifts. The Foundry treats every tool as a replaceable slot.

No Feedback = No Improvement

Most agent pipelines are fire-and-forget. Without evaluation loops, you never know what's working. The Foundry runs Karpathy loops on every slot to surface winners.

Manual Orchestration Doesn't Scale

Humans shouldn't wire APIs together by hand. The Foundry composes chains automatically from tournament-winning micro-agents, then executes them slot-by-slot.

From Typed Contracts to Autonomous Chains

Five stages transform a business intent into a self-optimizing autonomous workflow. No code. No vendor lock-in. Just contracts and competition.

01

Slots -- Define Typed Contracts

Each slot declares a typed contract: what goes in, what comes out, and which KPIs measure success. Think of it as an interface that any micro-agent can implement. Input schema, output schema, evaluation criteria -- all explicit.

input → output + KPIs
02

Vault Scan -- Search 4,230 Skills

The system scans the AaaS Vault -- 4,230 skills across 52 categories -- to find every possible approach for each slot. Firecrawl vs. Apify for scraping. Resend vs. SendGrid for email. All candidates surface automatically.

52 categories · 4,230+ skills
03

Tournament -- Karpathy Loop Picks Winners

Each candidate micro-agent competes. The Karpathy evaluation loop runs them against real data, scores results on the slot's KPIs, and assigns Elo ratings. The best approach rises to the top. Losers get archived.

Elo rating · automated evaluation
04

Chain Compose -- Wire Champions into Workflows

Tournament winners from each slot are wired together into complete chains. The email chain gets the best lead finder, the best enricher, the best writer, the best sender -- each independently proven.

champion → champion → champion
05

Execute -- No Human in the Loop

Chains run slot-by-slot, fully autonomous. Each slot receives the previous slot's typed output, processes it, and passes structured results downstream. Monitoring and metrics are built into every execution.

autonomous · slot-by-slot

System Status

Current production metrics from the Micro-Agent Foundry.

49
Slots Defined
664
Micro-Agents Generated
50
Tournaments Completed
15
Chain Families Active
77
API Secrets Available
111
Cloud Functions Deployed

15 Autonomous Chains

Each chain family is a complete business workflow decomposed into typed slots. Tournament winners fill every position.

Email Outreach

6 slots

Content Marketing

3 slots

Competitor Intelligence

3 slots

Lead Qualification

3 slots

Customer Onboarding

2 slots

Sales Proposal

2 slots

SEO Audit

2 slots

HR Recruitment

2 slots

Financial Analysis

2 slots

Product Feedback

2 slots

Brand Voice

1 slot

Deep Competitive Research

5 slots

Social Media Engine

4 slots

Website Optimization

5 slots

Context Framework

3 slots

Email Outreach Chain

The most complete chain: 6 slots, each filled by a tournament-proven micro-agent. Here is the full pipeline with the real APIs powering each slot.

01

Lead Finder

Discover target companies and contacts from web scraping and search APIs

Firecrawl Brave Search Apify
02

Lead Enricher

Crawl company domains to extract context, tech stack, and decision-maker info

Firecrawl crawlDomain()
03

Sequence Writer

Generate personalized multi-step email sequences from enriched lead context

Vertex AI
04

Send Executor

Deliver emails with tracking, unsubscribe headers, and rate limiting

Resend API
05

Metric Collector

Track opens, clicks, replies, bounces, and unsubscribes per sequence

Firestore email_logs
06

Evaluator

Karpathy loop scores sequence performance and triggers re-tournament if below threshold

Vertex AI Karpathy Loop

How to Trigger

Three ways to activate the Foundry. Each triggers a full chain execution with tournament-proven micro-agents.

Email Trigger

// Send to any chain family
TO: email-chain@aaas.select
TO: seo-chain@aaas.select
TO: content-chain@aaas.select

// Format:
{family}-chain@aaas.select

API Trigger

// POST to foundry endpoint
POST /foundryMegaGrind

{
  "chain": "email",
  "target": "saas-companies",
  "limit": 50
}

Schedule

// Automated cadence

Weekly   Vault scan
          New skills → new candidates

Monthly  Full re-grind
          All slots re-tournament