FUURAA AI Knowledge Library · Record check tool

AI Agent Commissioning Decision Record Check Tool

Run ten structural, evidence and temporal checks on a commissioning decision JSON locally in your browser. The file is not uploaded to FUURAA.

Tool statusPublicly usable · no upload by defaultCheck rulefuuraa.ai-agent-commissioning-decision-record/v1Sources checked17 August 2026

Record check workspace

Load a commissioning decision record, then close ten bounded checks.

Choose a local JSON file, paste content, or load a blank template or clearly labelled fictional demonstration. The check reads only text in this page.

Privacy boundary: this tool does not submit, store or transmit the selected file. Continue to handle sensitive or restricted material under your organisation's rules.

Applicability boundary

This tool checks whether a record is reviewable—not whether an agent is safe or releasable.

It can check

  • JSON, v1 rule identity and the fifteen-field structure
  • Whether release, use, evidence, authority, recovery, canary, monitoring and incident fields are missing
  • Whether decision, separated owners, expiry and reopen triggers are present

It cannot prove

  • That inventory is complete, evidence authentic, rights cleared or no unknown dependency exists
  • That evaluation is sufficient, controls effective, rollback viable or the canary representative
  • That the agent is safe, release justified or legal, audit and certification requirements met

FUURAA analysisMachine checking should first block records that cannot bind exact release, excluded uses, real authority, open findings, recovery, canary thresholds, emergency revocation and an expiring decision. After all ten pass, independent reviewers must still open source evidence, repeat consequential tests and confirm that the actual deployment exactly matches the record.

Standards and method sources

Structural standards can check record shape; commissioning still depends on source evidence.

Internet-Draft · 16 June 2022

JSON Schema 2020-12 · Validation Vocabulary

Provides concepts for required properties, types and focused structural validation.

BoundaryThis tool implements one focused FUURAA v1 rule, not general JSON Schema.

Open primary source ↗
Internet Standard · December 2017

IETF RFC 8259 · JSON

Defines the portable data format parsed locally in the browser.

BoundaryValid JSON cannot prove that evidence is authentic, controls work or commissioning is justified.

Open primary source ↗
Published 26 January 2023

NIST AI RMF 1.0

Connects governance, context, measurement and risk treatment across the AI lifecycle.

BoundaryVoluntary and use-case agnostic; it does not authorise a release or certify a deployment.

Open primary source ↗
Final published 26 July 2024

NIST SP 800-218A

Adds AI-specific secure-development practices for model producers, system producers and acquirers.

BoundarySecure-development guidance does not prove operating safety, control effectiveness or release suitability.

Open primary source ↗
Published 27 November 2023

UK NCSC · Secure deployment

Covers infrastructure and model protection, incident preparation, responsible release and usable guidance.

BoundaryHigh-level security guidance does not evaluate a particular agent, environment or decision.

Open primary source ↗
Published 27 November 2023

UK NCSC · Secure operation and maintenance

Connects behaviour and input monitoring, secure updates and lessons learned to deployed operation.

BoundaryIt does not set task-specific thresholds, sector duties or universal retention periods.

Open primary source ↗
Draft EN dated September 2025

ETSI EN 304 223 V2.0.0

Provides baseline cybersecurity requirements from design and deployment through maintenance and end of life.

BoundaryA baseline security standard is not a complete safety, privacy, sectoral or commissioning decision.

Open primary source ↗
W3C Recommendation · 30 April 2013

W3C PROV-DM

Models entities, activities, agents, time, derivation and responsibility for provenance.

BoundaryProvenance shows lineage; it does not prove truth, rights clearance or release suitability.

Open primary source ↗

Continue the evidence chain

Return to the record template and commissioning protocol, then hand an authorised decision into operating evidence.

Build a commissioning decision recordRead commissioning and authority handoverVerify an Agent evaluation recordBuild an Agent operating-period recordEnter AI Evidence AtlasReturn to AI Knowledge Library