REPOSITORY / 01Python
readable-mcp
Safe URL-to-Markdown extraction for AI clients.
View source on GitHub ↗Open source / Public proof
The implementation can be inspected.
These repositories expose the patterns behind the service: controlled network access, durable work, explicit orchestration, and complete product surfaces.
ENGAGEMENT STATUS
- Delivery
- Senior-led
- Scope
- Fixed boundary
- Handoff
- Documented
- Location
- Remote / EU
REPOSITORY / 02Python
mcp-job-queue
Durable background work for long-running MCP tools.
View source on GitHub ↗REPOSITORY / 03TypeScript
agent-orchestrator
Explicit coordination patterns for bounded agent work.
View source on GitHub ↗REPOSITORY / 04TypeScript
ssrf-safe-fetch
Reusable network safety controls for server-side fetching.
View source on GitHub ↗REPOSITORY / 05TypeScript
fcm-push-service
Push notification infrastructure for agent completion events.
View source on GitHub ↗REPOSITORY / 06Kotlin + TypeScript
tether
A mobile control surface for Claude Code.
View source on GitHub ↗How to read the proof
Look past the happy path.
Inspect how inputs are validated, how network destinations and permissions are bounded, what happens after a timeout, and whether errors are legible to both people and AI clients.
A repository is not a production guarantee. It is evidence of engineering judgment—and a concrete starting point for technical due diligence.
Paid entry engagement · €1,500
Start with the system boundary.
The Architecture Review turns a use case, prototype, or delivery opportunity into a defensible implementation plan in five business days.