projects
Tooling, instrumentation, and concept work for AI-native research.
A common thread runs through these: the runtime that consumes a tool can produce evidence about it. Hooks first. Configuration follows. Research stops being a separate motion and becomes a continuous read of a live system.
Cleo Smart Skills
shipped · may 2026Researcher + builder
An evidence-driven skill catalog with a learning loop wrapped around it. Built in 3 days at a hackathon, hand-written code under 100 lines. Two MCPs, project hooks across Claude Code and Cursor, threading + clustering backend, curator UI.
Internal Slack Helper
in productionResearcher + builder
A Slack-native AI helper running a local model with MCP connections to internal data sources. Every interaction becomes evidence; the configuration is the product, iterated continuously based on usage traces.
Inscryption Agent
in progressPersonal
Personal project: an agent that plays the deckbuilder Inscryption. Exploring agent design, decision policies under partial information, and self-improvement through play traces.
Intersectional Decision-Support Tool
concept · scopingPersonal
Concept: research-backed tool providing situated guidance based on user context. Drawing on social-anthropological framing of how identity, environment and incentive structures shape decisions.