Namaste Ji by Ayushman Dash

How the tools connect — and how this repo is itself built by agents.

The dev loop is a circuit: a PR opens, CI checks, merge deploys exactly what changed, Alchemy keeps the substrate honest, every model call goes through the Gateway, every prompt and trace lands in Langfuse, and the coding agents that wrote most of this reach those systems through MCP — with the founder reviewing strategy and merging. The same shape the product runs on.

The circuit.

push / PRgreen → maininfra.ymlwrangler deploy · pages deployprovisionsevery model calltraces · promptsrun eventsPRstoolsdev loopprompt edits via MCP GitHub repomonorepo · PRs only · CODEOWNERSCI (Actions)typecheck · tests · secret scan · JSONdeploy-*.ymlpath-scoped · merge→staging · prod gatedAlchemy IaCinfra/alchemy.run.ts · stage pinnedstaging on Cloudflare7 Workers · 3 Pages · D1/R2/KV/Q/VecAI GatewayBYOK OpenAI · Cohere · Sarvam · Workers AILangfuseprompts (name@label) · tracesMCP serversGitHub · Cloudflare · Langfuse · Notion · HFcoding agentsClaude Code + the founder in the looplocal meshdocker compose · wrangler dev · local LLMconsoleruns · approvals · Control DO · KV

The loop, in words.

1 · Branch → PR → green → squash-merge

Conventional commits; small single-purpose PRs; CI = typecheck every workspace, unit tests, a secret-scan that fails the build if any secret-bearing file is tracked, JSON validation. main is protected by convention (free plan) and by a prepared ruleset.

2 · Deploy exactly what changed

Path-scoped workflows: deploy-console, deploy-<agent>, deploy-showcase… each runs migrations before code, deploys with wrangler, and smoke-checks the result (the console's checks every JS chunk; this site's checks every referenced asset).

3 · Infra as code, account-pinned

Alchemy provisions D1 · KV · R2 · Queues · Vectorize · AI Gateway · Secrets Store per environment, hard-pinned to the canonical account and stage (a near-miss taught us). Bindings are declared per env — never inherited.

4 · One gateway, one observability stream

Every model call → AI Gateway (BYOK per provider, per-env tokens); every prompt → Langfuse by name@label; every run → events the console can show. A model swap is a KV edit; a prompt change is a new version with a label, never auto-promoted to production.

5 · Agents build it, the founder steers

AGENTS.md is the front door for coding agents (repo map, stack, conventions, the MCP servers to prefer). Claude Code opened most PRs on this repo; the founder reviews strategy, design decisions and merges. Memory notes carry gotchas between sessions.

6 · Runs without the cloud

docker compose brings up console-api + console + the agents with mock identity; a native llama.cpp serves a local LLM; the Lab works Vectorize-free (raw-embedding copy in D1). Cloud providers are reachable locally through the Gateway (pause WARP first — a real gotcha).

Go deeper

In the repo

  • .github/workflows/*.yml
  • infra/alchemy.run.ts
  • docker-compose.yml
  • AGENTS.md · CLAUDE.md (the agent front door)
  • .mcp.json (gitignored — tokens inline)