Everything we've published, in one ledger.
Each entry below is a maintained, versioned artifact — not a landing page. Specifications carry version numbers; submissions carry their full text. Additional research briefs are forthcoming and will be listed here as they publish.
Published artifacts
Decision Receipt Specification →
The receipt format: structure, deterministic evidence IDs, validation rules, and trust model for machine-verifiable decision provenance.
Cognitive Security Framework →
Seven functions from IDENTIFY to ADMIT, assurance levels CAL-1 through CAL-7, a five-level maturity model, and paste-ready procurement language.
AGD-Bench Methodology →
The Verified Governed Delivery Score: eight task categories and GA thresholds for measuring whether agentic delivery can be proven, not just performed.
Decision Failure Atlas →
Documented machine-generated decision failures — each mapped to the governance control that would have prevented it.
Policy Submissions →
The NIST SP 800-218 Rev. 1 public comment and the NIST AI Consortium letter of intent, published in full.
Federal Capability Statement →
Entity details, HUBZone status, mission surfaces, and compliance posture for contracting offices and program managers.
Source text for the standards and submissions is maintained in the open repository — github.com/BrianCLong/summit.
More briefs are on the way.
The publication program follows the engineering program: when a capability ships and survives measurement, it gets written up. Watch the repository or the newsroom for new entries.