π London | ποΈ Chief Engineer at SHOTClubhouse | π€ Agentic engineering builder
I build practical tools for coding with agents: context engineering, long-term memory, workflow telemetry, and small CLIs that remove daily drag.
- ποΈ Chief Engineer at SHOTClubhouse - building toward an agentic operating system for growing SHOTClubhouse
- βοΈ Every useful tool, workflow, and agent loop gets folded back into that direction: product, club portals, content, ops, and delivery
- πΊ Wololo Platform - autonomous agentic engineering operating system work feeding into SHOTClubhouse: onboarding, provisioning, governance, observability, and agent-fleet operations
- πΆοΈ Cofounder + Tech at CoChilli - building out cochilli.co.uk and the commerce, ops, and delivery systems around it
- πΌ PopAJob - Tinder for jobs, currently closed source and planned to open later
- ποΈ TalkToMe - cross-platform AI chat and voice app under development, with credit to crmitchelmore/justspeaktoit for the original voice-to-text inspiration
- π§ JSTLFT - soon-to-open-source product experiment, kept intentionally brief until the public release
- βοΈ stevengonsalvez.com - writing on agents, automation, security, and practical software delivery
- π dev.to/stevengonsalvez - syndicated posts with canonical links back to the blog
- π¦ agents-in-a-box - context engineering for agentic coding
- π§ ainb-reflect-memory - long-term memory for AI coding agents
- π§° ainb-toolkit - curated AI coding skills and per-tool rule sets
- π°οΈ cerebro - local-first, token-minimal daily tech-intelligence pipeline
- π qstatus - Amazon Q Developer usage monitoring for macOS and CLI
- ποΈ promptregistry-mcp - lightweight file-based prompt server for developers
- π§© mcp-ethicalhacks - MCP security examples and failure modes
- βοΈ dotfiles - local development environment
- Building agent workspaces - Keeping prompts, skills, memories, telemetry, and code close enough to survive real work
- Compressing agent cost - Measuring token waste, trimming context, and routing judgment to the right model
- Writing about AI workflows - Publishing practical notes on stevengonsalvez.com
- Turning workflow pain into tools - Small CLIs, dashboards, and repo conventions before big platforms
- The Token Optimisation Playbook
- Opus vs GPT on Real Ops, Part 2
- The Underappreciation and Rebirth of Warp
- Opus vs GPT on Real Ops: Same Brain Food, Different Brains
- 2-2 Factor for AI Agents: Multi-Agent Reliability
Ship the smallest useful loop, measure where it breaks, then encode the lesson so the next agent run starts smarter.
Random Facts
- I like tools that make hidden workflow state visible.
- I prefer local-first systems until the network earns its place.
- I write docs because future me has terrible memory.
- I care more about reliable loops than fancy demos.




