Documentation Generator
Keeps architecture, dependency, and setup documentation current by analyzing the code directly, so docs don’t lag behind the codebase
In production
What governed AI looks like running live
Not a demo. Not a prototype. This is the same three-layer stack I build for clients, running against real engineering data: connectors feeding it, domain knowledge steering it, and specialist agents detecting patterns humans miss.
Layer 1
Connect
MCP connectors feed live data in (Git history, issue trackers, internal APIs) through governed, identity-aware access.
Layer 2
Teach
Domain knowledge and steering give the agents context: what the metrics mean, what the workflows are, what good looks like.
Layer 3
Orchestrate
Specialist agents activate on quantitative markers, each grounded in validated behavioral research, then synthesize findings into a narrative.
The agents doing the work
These are agents I’ve built and deployed. Not a wishlist. They handle codebase analysis, documentation, security, infrastructure, and operations work, and they’re what the orchestration layer routes to when it detects work that needs doing.
Showing 31 agents
ADR Generator
Creates Architecture Decision Records from significant changesets using git history and LLM synthesis
Git History Analyzer
Parses git log, groups commits into logical changesets, and identifies significant architectural changes
RAG Answer Agent
Answers questions from your knowledge base using retrieval-augmented generation with source citations
Onboarding Guide Agent
Turns architecture docs and codebase analysis into structured onboarding guides, so new engineers ramp up in days, not weeks
Changelog Agent
Transforms grouped changesets into user-facing release notes with categorized entries
Frontend Builder
Generates frontend components and pages from architecture specs and design requirements
DB Builder
Generates database schemas, migrations, and query layers from data model specifications
Firebase Builder
Scaffolds Firebase projects with Firestore rules, Cloud Functions, and hosting configuration
Lambda Builder
Generates AWS Lambda functions with proper handlers, IAM policies, and deployment configs
Frontend Agent
Analyzes frontend codebases for component patterns, UI frameworks, and rendering strategies
Test Agent
Generates unit and integration tests based on code analysis and coverage gaps
QA Agent
Reviews builder output against specs, triggers revision cycles with feedback until acceptance criteria pass
PR Review Agent
Reviews pull request diffs for correctness, performance issues, and adherence to team conventions
Architecture Agent
Maps architectural patterns, layer structures, auth strategies, and multi-tenancy in codebases
Data Discovery Agent
Scans codebases to identify BigQuery schemas, API endpoints, query patterns, and data gaps
Code Reusability Agent
Identifies reusable components, API patterns, utilities, and coding conventions across a codebase
Dependency Agent
Scans package manifests for outdated, vulnerable, or copyleft-licensed dependencies
Database Agent
Analyzes database schemas, relationships, indexes, and constraints from codebase artifacts
API Agent
Discovers API routes, HTTP methods, auth requirements, and OpenAPI specs across a codebase
Testing Analysis Agent
Evaluates test coverage, identifies untested modules, and reports code-to-test ratios
Cost Estimator Agent
Estimates monthly infrastructure spend based on detected cloud services, databases, and compute resources
Migration Planner Agent
Sequences architecture transitions with risk assessment and rollback strategies, so no one has to write the migration plan by hand
Performance Agent
Finds N+1 queries, missing indexes, unbounded loops, and pagination gaps across your codebase before they slow down users
Security Agent
Scans for hardcoded secrets, auth weaknesses, input validation gaps, and dependency vulnerabilities before they ship
Generation Security Agent
Validates AI-generated code for security risks before it reaches production
License Compliance Agent
Cross-references dependencies against license policies and produces pass/fail compliance reports
Digital Marketing Agent
Runs SEO audits, identifies content gaps, and drafts keyword strategies matched to your positioning
SEO Audit Agent
Ingests sitemaps and page HTML to surface missing meta tags, broken links, and Core Web Vitals issues
Report Synthesizer
Consolidates outputs from all analysis agents into structured markdown reports with severity-ranked issues and recommended next steps
DevOps Agent
Analyzes CI/CD pipelines, containerization, IaC tools, and environment configuration patterns
See how this applies to your team
Book a free call and I’ll map which layers of this stack would move the needle for your team.
Book your free callFree, zero-pressure, and you’ll leave with something useful either way.