Canvis — a personal agent system

Canvis is my personal agentic system: a long-running, autonomous agent built on Claude Code that operates as an extension of how I think and work, rather than a collection of single-task tools.

The bet is simple — most "AI productivity" stops at the chat window. Canvis is an attempt to go further: persistent context, its own memory, scheduled routines, and a governance layer that decides what the agent can do on its own versus what needs me in the loop.

What it does

• Research — scoped, cited research that persists as durable files, not throwaway answers.
• Content — drafts thinking on AI transformation grounded in things I've actually built.
• Strategy & ops — keeps a decision ledger, runs daily and weekly review routines, and surfaces the calls that need my judgement.

How it's built

The core idea is a template factory: every capability is generic and reusable, routed through a layered system — always-on context, modular rules, on-demand skills, deterministic hooks, and specialised sub-agents. The scarce input is my review attention, so the system is designed to spend it on judgement, not on catching errors.

More writing on the design and what I've learned is in the Writing section.