Quipu: A Governed Bitemporal Knowledge Graph Store

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Steve Brown
• Venue: arXiv: cs.AI
• Published: 2026-08-17
• Collected: 2026-08-18
• URL: http://arxiv.org/abs/2608.16813v1
• PDF: https://arxiv.org/pdf/2608.16813v1
• DOI: UNKNOWN
• Registry group: academic_apis_and_preprints
• Evidence role: canonical_academic_index
• Language: en
• Country / region: UNKNOWN

Plain-Language Brief

The paper presents Quipu, an embeddable knowledge graph store designed to handle agent workloads more effectively than existing stores. Quipu reverses four common defaults of existing stores: it only accepts facts through a gate that evaluates their post-state, uses bitemporal data and rules for trust management, allows named graphs as the unit of authority and trust, and embeds governance specifications in the store itself. The paper claims that Quipu significantly improves upon ungated stores in terms of accuracy, compliance with contractual obligations, and auditing. It also demonstrates Quipu's effectiveness through experiments using a deterministic multi-writer lifecycle called Census.

What the Document Says

• The paper introduces Quipu, an embeddable knowledge graph store.
• Quipu reverses four defaults of existing stores: accepting facts through a gate, using bitemporal data and rules, allowing named graphs as the unit of authority, and embedding governance specifications in the store.
• The paper claims that Quipu improves accuracy, compliance with contractual obligations, and auditing compared to ungated stores.
• The effectiveness of Quipu is demonstrated through experiments using a deterministic multi-writer lifecycle called Census.

Key Concepts Mentioned

• Quipu
• knowledge graph store
• bitemporal data
• trust labels
• verdicts
• governance specification
• Census
• deterministic multi-writer lifecycle

Claims Made by the Document

• Quipu improves accuracy compared to ungated stores.
• Quipu complies with contractual obligations better than ungated stores.
• Quipu provides effective auditing.

Evidence Used

• Results from experiments using Census

Methods / Approach

• Experiments using a deterministic multi-writer lifecycle called Census

Limitations Stated by the Authors

• UNKNOWN
• The effectiveness of Quipu may be specific to the experimental setup and may not generalize to other scenarios.

Limitations Noted by the Agent

Agent interpretation — not from the source.

• The evaluation is based on a single study. More research is needed to verify these findings in different contexts.

Open Questions

• How well does Quipu perform under various workloads and real-world conditions?
• What are the performance tradeoffs between Quipu's features?

Notes for Future Q&A

May include interpretation. Labeled where it does.

• The paper focuses on demonstrating Quipu's effectiveness, but it does not discuss potential drawbacks or limitations in detail.

Raw Abstract / Excerpt

Agents now write knowledge graphs, but knowledge-graph stores still carry defaults set when humans curated them: accept writes now and clean later, keep one time axis or none, treat every writer's facts as equally trustworthy, and leave governance to dashboards and middleware. These four defaults are individually convenient and jointly untenable under agent workloads. We present Quipu, an embeddable store that inverts all four: no fact enters except through a gate whose predicates evaluate the pending post-state; data, trust labels, verdicts, and the rules themselves are bitemporal; named graphs are the unit of authority and trust, composed under a lattice whose one invariant is that composition never widens; and the governance specification $Σ$, the trace, and signed verdicts are facts in the store they govern, making the audit $T \models Σ$ a query. We evaluate with Census, a deterministic multi-writer lifecycle whose single seeded run scores every research question against planted ground truth: the gated store ends with 0 of 6 planted defects versus 6 of 6 ungated; all 7 composition probes uphold the lattice contract; 50 of 50 satisfied verdicts re-derive faithfully as of their instant while all 50 would be misreported under a latest-only rule set; and the SARC reference checker agrees with the in-store audit verdict-for-verdict, differing only on coverage semantics. A recorded trace from a governed writer surfaces a live enforcement gap the audit names with its remediation. On DEMM-Bench, an external decision-evidence sufficiency benchmark, a content-only reading of the exported records answers all 512 property-level governance questions correctly with zero overclaim under all eight degradation conditions, while container-presence baselines overclaim on up to 87.5% of them -- and the run surfaced, and led us to close, a gap in what a denial's verdict attests.