Products - TinCan Labs

Core platform

OpenArchitect — the AI-native control plane

Coding agents are the harness for AI coding. OpenArchitect is the harness that sits upstream: we turn enterprise intent into governed requirements, architecture, and executable delivery paths, then route implementation through downstream agents and enterprise delivery systems. The result is a governed loop from strategy to production change to measured outcomes and back into the next cycle.

	Coding Agents	OpenArchitect
Context	Repository code, file structure, dependencies	Enterprise context (vision, strategy, teams, systems, constraints), curated vertical standards, and repositories across the architecture cascade
Constraints	Language rules, test suites, linting	Architecture guardrails, governance policies, regulatory requirements, domain boundaries
Domain knowledge	Language docs, framework APIs	Curated vertical industry standards, partner dialect adapters
Governance	Code review, sandbox isolation	Full audit ledger, constraint enforcement, traceability from intent to implementation
Output	Working code in a branch	Architecture decomposition, delivery paths, governed implementation targets
Loop closure	Tests and code review inform the next code change	Production outcomes, metrics, and operator feedback inform the next strategy and architecture cycle

Business model

Open core distribution, proprietary intelligence

The protocol wants distribution. The proprietary data layer wants monetization.

Free tier

Open-source protocol and runtime
Reference implementation
ENTERPRISE.md standard
Community support

Value tier

Curated industry knowledge graphs
Pre-built architecture skills
Enterprise graph hosting
MCP tools and integrations

Enterprise tier

Partner dialect adapters
Custom constraint libraries
Managed platform with SLA
Audit, compliance, and governance features

Product portfolio

Building blocks

ENTERPRISE.md

Open source

A proposed standard that enables AI agents to automatically navigate enterprise-scale, multi-repository environments through progressive disclosure. It extends the familiar AGENTS.md pattern with level-aware entrypoints, deterministic routing catalogs, and governance-aware traversal.

Progressive disclosure — Move from enterprise intent to implementation context without loading everything at once
Level-aware entrypoints — ENTERPRISE.md, SOLUTION.md, and DOMAIN.md at the right architectural altitude
Deterministic routing — Canonical catalogs such as initiatives.yml, domain-workstreams.yml, and domain-implementations.yml
Governance-aware automation — Fail closed on missing selectors, ambiguous routes, and inactive states
Git-native adoption — Start with files, templates, and CI validation rather than a new platform

Explore the spec

Security Review SKILL

Self-service

A self-service security review solution that integrates directly with the IDEs your team already uses. Works with the LLM models you already pay for, providing deterministic reasoning guidance with industry-standard security knowledge frameworks.

IDE integration — Works directly in your existing development environment
Use your existing LLMs — No additional AI subscription needed
Deterministic reasoning — Structured guidance for consistent, reproducible security reviews
Industry knowledge injection — OWASP ASVS, NIST SP 800-53, and other security frameworks
Expert review option — Professional security assessment available for complex scenarios

TMF Knowledge Graph

Platform

A curated, AI-usable knowledge graph built from TMF standards. Anyone can download the raw specs; nobody else has done the curation, linking, and embedding work to make them machine-operational for AI-assisted architecture.

2,846 eTOM processes — Curated, linked, and searchable
1,744 SID data entities — Full information framework coverage
1,308 Functional Framework functions — Mapped to architecture workflows
27 ODA components — Open Digital Architecture alignment
25+ TMF MCP tools — Exposed for agent consumption

Integration

Get started in minutes

Add the TinCan Labs MCP server to your IDE configuration and drop the SKILL folder into your skills directory.

Step 1: Configure MCP server in your IDE settings
{
  "mcpServers": {
    "sec-mcp": {
      "url": "https://url_will_be_provided/mcp",
      "transport": "http",
      "headers": {
        "API_KEY":"your_tincan_token"
      }
    }
  }
}

Step 2: Drop the SKILL folder
.claude/skills/security-review/
or in
.codex/skills/security-review/

Your IDE now has access to deterministic security reasoning powered by your existing LLM models.

Get started

Interested in OpenArchitect?

We work with enterprises that are serious about compressing the time from strategic decision to production change. If that sounds like your organization, we would like to hear from you.

Learn about our approach Read our thinking