GoKawiilOpen-source author proposes shared memory spec for AI coding agents
A developer behind the open-source tool LoreKit argues that AI coding assistants currently store 'memory' in incompatible formats, including AGENTS.md files, Cursor's .mdc rules, Anthropic's memory tool directories, GitHub Copilot's repo facts, and various MCP-based knowledge graphs. The piece surveys these formats and existing proposals, drawing a parallel to observability tooling before OpenTelemetry unified it with a common data model.
GoKawiil's interpretation of the reporting above, not reported fact.
The lack of a shared memory format means agent-learned context is siloed per tool, so switching or combining AI coding assistants means losing accumulated context, which could slow adoption of multi-agent workflows. The comparison to OpenTelemetry suggests standardization could unlock interoperability, though the author acknowledges LoreKit's own format is just one of several competing proposals rather than a ready-made solution.
- AI coding tools currently use at least four incompatible memory storage formats
- The fragmentation mirrors pre-OpenTelemetry observability tooling, per the author's analogy
- LoreKit's creator is calling for community agreement on a shared memory spec, not promoting LoreKit itself as the fix
Source: lorekit.io — Mads Thines, 2026-11-08
Published there as: “Agent memory needs its OpenTelemetry moment”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.