AI DevKit

AI DevKit gives a team running several AI coding agents one local-first operating layer instead of a separate, disconnected setup per tool. A single configuration file reconciles setup across the coding agents in use, a live terminal dashboard supervises running agent sessions across providers, and a command lets a developer route prompts, logs or test output to an already-running agent.

Its memory component stores decisions, conventions and fixes in a local SQLite database that agents search only when needed, rather than carrying the whole history in every prompt, exposed through both the Model Context Protocol and the CLI. A set of composable skills, led by a project's development-lifecycle skill, combine with verification, test-driven development, review, debugging, security, documentation and simplification skills into repeatable workflows. It installs with a single command and is aimed at teams whose AI coding setup has grown into several agents with no shared control surface or memory of yesterday's conventions.