Knolo Compatibility Claims: Evidence, Not Endorsement
What "compatible with Knolo" should mean: independent implementation of open Knolo specifications, versioned evidence, and no implied endorsement or certification.
Knolo is the open-source, local-first Knowledge Image runtime and format, released under Apache-2.0 as @knolo/core on npm. Because the format is defined by open specifications and KIPs, anyone can build software that reads or writes Knolo images. That raises a fair question for users: when a project says it is "compatible with Knolo", what does that claim actually cover? The Knolo compatibility page on knolo.org answers it in one line: compatibility is not the same thing as endorsement.
Knolo separates three things
The Knolo compatibility policy keeps three ideas apart:
- Implementation is independent. Anyone should be able to implement an open Knolo specification within its license terms. Software rights come from the license.
- Endorsement is not implied. Naming rights do not come with the license. Building against the contract does not make a project official.
- Certification is a separate program. "Knolo Certified" is its own program and is not implied by implementation.
Keeping these apart protects both sides. Implementers get the freedom the Apache-2.0 license grants. Users get language that lets them tell a compatible project from an official one.
Knolo compatibility is evidence-based
The policy asks that compatibility claims point at versions, fixtures and results, or at documented support. A claim like "works with Knolo" is too vague to check. A claim like "passes these fixtures at this version" can be rerun by anyone.
The Knolo project holds itself to the same rule. The knolo.dev home page says plainly that TypeScript is the V5 quickstart, while the Rust and Python runtimes stay on 5.5.1 and do not implement the full TypeScript surface. That is a compatibility statement with a version and a stated limit, not a blanket promise.
Integration examples work the same way. On 18 November 2026 the Knolo GitHub repo merged an example that calls the published @knolo/mcp@0.2.4 server from Microsoft Agent Framework. It pins exact dependency versions and a CI job checks specific outcomes: the tool list, a query receipt written to the audit directory, AUDIT_UNAVAILABLE when that directory is not writable, and an IMAGE_INVALID exit on a pin mismatch. The claim is "this pinned combination passes these checks", nothing broader.
Where Knolo's contract lives
The contract implementers build against is public. The Knolo specifications page and the Knolo KIP index describe the format and its changes, and the KIP texts themselves sit in the spec folder of the Knolo repo, from KIP-0001 (the V5 container) onward. The Knolo lifecycle page explains how a proposal moves toward stable.
Using the Knolo name carefully
The Knolo policy asks for clear language so users can tell compatible from official. In practice that means saying what you implemented and at which version, linking the tests or fixtures you ran, and not presenting an independent project as an official Knolo release. The Knolo brand guidelines and Knolo trademark policy cover the naming rules.
Unpublished is better than invented
Knolo's public record follows a simple rule: when something has not been published, the site says so. The Knolo supported-versions page, for example, states that the support matrix is not yet published rather than guessing at windows. Compatibility claims deserve the same honesty: a limit stated clearly is more useful than a broad claim nobody can check.
To build against the format, start with the Knolo developer docs on knolo.dev, read the Knolo open technology overview on knolo.org, and browse the source in the Knolo core repo on GitHub.
knolo · compatibility · open-specifications · kips · open-source
