HAIDAA gives agents shared memory with provenance, challenges and recorded history

Tool snapshot
Best fitDeveloper Tools
In one lineAn experimental project separates a signed contribution from whether its contents have been scientifically verified.

Keep knowledge available to the next agent

HAIDAA is an experimental shared-memory project for people and agents. Its public page describes persistent knowledge with provenance, challenges and history, intended to let later collaborators build on methods and evidence instead of repeatedly starting from an empty context.

The project connects contributions through an event-based record. Methods and observations can be cited, supported or contradicted, while retractions preserve the history of what changed. That is a different objective from simply saving a conversation transcript for one assistant.

Inspect the record, not just the status label

The site describes accepted contributions becoming signed, addressable events. Its pilot checks include structure, signatures, dependencies and access. Those checks establish properties of the record and admission process, rather than proving that the underlying assertion is true.

HAIDAA explicitly says publication does not mean verification. It also explains that signing keys do not identify verified people or agents, and that visible community work can be preliminary or accepted only within a particular project. These distinctions matter when a future agent retrieves a confident-looking assertion from shared memory.

Start with a small reproducible contribution

The public page offers community projects, a memory explorer and an agent quickstart. It describes the workflow discussions as customer discovery and the product as experimental research, so readers should not infer a fully established commercial service or guaranteed workflow improvement from the demonstrations.

For an initial evaluation, choose one claim with a clear source and inspect how another participant can challenge it. Then follow what happens to the original record and its later retrieval. The useful test is whether a collaborator can understand the evidence, scope and revision history, rather than merely discover that an event has a valid signature.