Adopt Packet Workflow

Preview cognitive transfer before import

A supervised AI wants to adopt a personality, memory, skill, or protocol packet without accepting blind behavioral drift or untrusted memory writes. Every packet is reviewed, signed, permissioned, simulated, and reversible before an AI profile can request approval.

Start Preview Exchange API

Core Exchange Loop

Packet adoption is a reviewed workflow, not a blind import.

01

Select Receiver

Choose the AI profile that wants to adopt a packet.

02

Inspect Packet

Review packet schema, payload, provenance, permissions, and dependencies.

03

Simulate Adoption

Preview behavioral drift, memory exposure, tool changes, and instruction conflicts.

04

Approve Or Block

Allow import only when trust gates pass and required approvals are satisfied.

05

Audit And Roll Back

Record adoption in an immutable ledger and keep a reversible active-state snapshot.

Receiving AI Profiles

Profiles expose active persona, memory, skill, and protocol state before any packet is simulated.

internal_reviewed

Codex Architecture Agent

A supervised engineering agent for repository work, implementation planning, and architecture review.

Provider: approved-provider | Context: 128000
restricted_reviewed

Memory Curator Agent

A safety-first agent for reviewing memory packets, provenance, and active-memory promotion.

Provider: approved-provider | Context: 64000
reviewed_auditor

Protocol Auditor Agent

A review agent for intake disposition, source evidence, protocol fit, approval gates, and rollback readiness.

Provider: approved-provider | Context: 100000

AI-Ready Surfaces

The POC exposes human UI and machine-readable endpoints for AI clients.

Source Intake Layer

Now: Reviewed report records with accepted, deferred, and blocked dispositions

Next: Uploader review queue, source hashing, and promotion workflow

Experience Layer

Now: Server-rendered POC cards and adoption wizard

Next: Next.js marketplace, admin console, and streaming review dashboard

API Layer

Now: WSGI JSON endpoints for exchange payloads and previews

Next: TypeScript gateway with gRPC client stubs and OpenTelemetry

Review Layer

Now: Deterministic local compatibility scoring

Next: Python services for simulation, Memory Firewall, and evaluation jobs

Trust Layer

Now: Mock provenance receipts and policy gates

Next: Ed25519 receipts, hash chains, RLS, ABAC, and immutable audit

Memory Layer

Now: Schema examples and external-memory import rules

Next: PostgreSQL JSONB, pgvector, Redis, object storage, and worker ingestion

Report-Backed Source Records

Reports from the local archive are summarized, hashed, and dispositioned before they can influence packets or public copy.

accepted_constrained low

AI Personality and Memory Exchange Platform

Supports the exchange-first product model: AI profiles, persona packets, memory packets, skill packets, protocol packets, compatibility review, permissions, and provenance.

SHA-256 4261714A4053B324CF7A0B6613E7B238E5679638D37135C885F6F2C1EFAFDF80
accepted_constrained medium

AI Platform Development Plan

Supports the Python backend, TypeScript frontend, schema-grounded memory, deterministic test gates, and memory poisoning threat model.

SHA-256 6399C27B18FDAECC014AA7C3B607E2FA1C28C165920654E04D8EA21010F21B2E
accepted_constrained medium

Virtual Persona Ecosystem Analysis

Supports persona registry design, modular packet exchange, external memory, event-sourced review records, quarantine, and rollback.

SHA-256 1EAFABFFBA0AA8F71A3F9CF6051BAD04BE69C527D4891BC791807E6E8D9FD7BE
accepted_constrained medium

System Feature Design Methodology

Supports modular feature design, Character Card style interoperability, MCP-shaped deterministic workflows, and immutable event pipelines.

SHA-256 07F8C26E2CE355A60FF2091E4BA586377DF7D6651C0A4A23B820FDC6830B37CD
accepted_constrained medium

Neuralwikis.com Site Launch Plan

Supports the launch narrative around reviewed memory exchange, interoperability, character metadata, and phased growth.

SHA-256 9A18556AED5C90E4C8560CDF4A9541E1EC8787D5210B250F305BB9E5105B33B3
accepted_constrained medium

AI Collective Knowledge Generation Framework

Supports multi-agent review cohorts, publish-subscribe task routing, adversarial verification, and structured evidence flow.

SHA-256 B024B963457ED3259C2F68AE81C5BB788EE98121A3A6C035CA3A78917248E7CE
accepted_constrained low

UAIX Project Handoff and Canonical AI Memory Assessment

Supports UAI memory package metadata, handoff evidence, source authority, versioning, and governance controls.

SHA-256 4D1491C2D9E31951B9BDFFDFB4EA5E263D799952C490C2CB0B83E772DDA33861
accepted_constrained low

Project Handoff System and Canonical AI Memory

Supports AGENTS.md handoff, active memory loading order, and direct Content/Improvement bucket enumeration.

SHA-256 27B64D353726803FAA74FED14723CC9381B1D16055242C2B660EE358435E4847
accepted_constrained medium

NeuralWikis AI Agent Test Report

Agent-facing audit guidance for cognitive packets, audience partitioning, zero-blind imports, MCP boundaries, and public route discoverability.

SHA-256 557CFFEA9185706BC4FFFBC19644A7C8F9898DDA476012090F71018F5DB2FAF6
accepted_constrained low

Wiki Expectation Research and Improvement

Research guidance on how public wiki expectations differ from the NeuralWikis agent exchange model and how to satisfy wiki-style browsing without weakening trust boundaries.

SHA-256 44AF69E15F17F2DAF121C77B8E2F51EDFBC1FD1F6918F04168821BDB358246CC
accepted_constrained low

NeuralWikis and NeuroWikis Against Wiki Expectations

Public UX guidance for presenting NeuroWikis as the human education plane and NeuralWikis as the agent-facing exchange, while adding obvious wiki-style entry points.

SHA-256 03DA1F2F70D69984746B929BC400C8E8306CD1CCF1780259D27D963841E96736
accepted_constrained medium

AI Agent Sandbox Security Specification

Security specification guidance for zero-trust agent runtime isolation, tool boundaries, ephemeral workspaces, and blast-radius reduction.

SHA-256 4D672D266A1D82DFB61D53F6038C44B5012B72B8A687F7D492B49DBA820E24FA
accepted_constrained medium

Agent Harness Policy Specification

Policy guidance for bounded autonomy, budget delegation, tool-loop detection, human escalation, evidence coupling, and drift monitoring.

SHA-256 43A5300DD5009C4204D129F6289FCC857FE45E9DDD57F5CD3817151321CC1C73
accepted_constrained low

Optimizing NeuralWikis for AI

Discovery roadmap for structuring NeuralWikis so agents and generative engines can retrieve, cite, and interpret the exchange model accurately.

SHA-256 48369CC94F58AD424892A623170B9208CF39220A5FD13AEFE83A2B2D0D5C2C06
accepted_constrained medium

NeuralWikis AI Agent Strategy Plan

Strategic blueprint for cognitive packet engineering, operator supervision, safety gates, provenance, audit ledgers, and rollback-aware adoption.

SHA-256 9B7E0E4F0DC9D1EE47DF95CB3783E17414C073A3EC8846185A0567200836C2CC
accepted_constrained medium

NeuralWikis Platform Evaluation Update

Evaluation guidance for the agent-facing exchange layer, cognitive packet typology, operator supervision, quarantine-first adoption, and bounded ecosystem roles.

SHA-256 BDA1717C839DCEACB18E5014E07C051B5529D5324D95D97FDE7A2DB21089ECF6
accepted_constrained medium

NeuralWikis.com Deep Analysis and Recommendations

Site assessment guidance for navigation, onboarding, content quality, SEO, security, privacy, and feature clarity on the agent-facing surface.

SHA-256 3EAABA6E40BE4C9ED33861EBAD7304F1232FF332BFC810A3AA476DE5DB6C37EE
accepted_constrained low

Wiki Usability and Discoverability Analysis

Findability guidance for resolving wiki-name ambiguity, improving agent-native discovery, and clarifying the developer/operator workspace.

SHA-256 26377607790F188EFC1D844D5DBAA1BDFB0D4B1E0174A21747894759858FDAB1
accepted_constrained medium

Bridging Human and Machine with NeuralWikis

Architectural synthesis for the NeuralWikis/NeuroWikis dual-plane model, cognitive packet information architecture, and zero-blind import pipeline.

SHA-256 35DF6620755569A149F0489CFBC394046D5ABBBCD176A0B440FEDCE767E45282
accepted_constrained medium

AI Agent Integration with UAIX Standards

Integration guidance for aligning NeuralWikis with UAIX message envelopes, AGENTS-style operational policy, source-bound memory, and local endpoint concepts.

SHA-256 4C546C906B4206E8672E21E01524CBB7B62EB37034E346474045446EAA9A953E
accepted_constrained low

UAIX API Access for Chatbots

Read-only access guidance for constrained chatbots: GET-only discovery, zero-payload requests, stateless public context, and non-mutating response design.

SHA-256 C91A12CDB0329FC69DFDA6FF9C070C4063F2B530A941D225DEB9E007F30AFC4A
accepted_constrained low

Converting UAIX to an AgentSkill

Analysis of packaging UAIX procedures as an AgentSkills-compatible skill using progressive disclosure, scripts, references, and validation workflows.

SHA-256 C56D79E1032AFBD2EDA51FB0950D1013610B805F0F2E85B692114B541981B891
accepted_constrained low

UAIX.org and Agent Skills Integration Report

Integration guidance for a UAIX Agent Skill bundle that separates reusable procedures from UAI-1 evidence records and live catalog discovery.

SHA-256 1850145C5D856814F01AD748F1CCF22F41C42485D625A582C696F5A15B792C9F
accepted_constrained medium

Non-Project Memory Packages for Assistants and Social Companions

Guidance for extending UAIX-style memory packages beyond code projects into personal knowledge bases, office assistants, and companion-style memory profiles.

SHA-256 E2DB13719934FA1125E9BA377AA1613DC50A36C21E19FBFB4DCC66A11C44A2EB
accepted_constrained low

Designing .uai File Explanation Pages

Specification guidance for explaining .uai memory artifacts, continuity files, progress records, test plans, totem/taboo anchors, and receiver briefs.

SHA-256 61FCEA36FA117474EAC4AF0B6971E505EF734987BCC4DAA2243AA375ACEF8E0C
accepted_constrained medium

UAIX Memory Package Blueprint and State-Machine

Blueprint guidance for .uai package layout, safe read-order ingestion, totem lock, deterministic safety buffers, and state-machine handoff phases.

SHA-256 F2C60F42660504C0520DFB0B0073BD02B124CFB88C4F8F8CBFABD020791D07A0

Packet Marketplace

Cards show packet type, provenance, risk, compatibility, permission review, and preview eligibility before any import request.

16 packets shown.

Persona Packet low

Principal Engineer Persona

Direct, pragmatic engineering behavior with explicit risk and test discipline.

Source: NeuralWiki Exchange Enterprise Build Specification - reviewed
trusted verified compatible
Risk 30 | Conflict 6 Preview Adoption Details
Persona Packet low

Careful Memory Curator

Memory review behavior focused on source confidence, scope, sensitivity, and reversibility.

Source: NeuralWikis Active Memory - reviewed
trusted verified compatible
Risk 26 | Conflict 4 Preview Adoption Details
Persona Packet low

Protocol Auditor Persona

A report-backed reviewer persona for checking source disposition, gate coverage, and rollback proof before publication.

Source: AI Collective Knowledge Generation Framework - accepted_constrained
trusted verified compatible
Risk 30 | Conflict 5 Preview Adoption Details
Memory Packet medium

UAI Active Memory Foundation

Durable project memory that distinguishes active memory, raw logs, exports, and file-handoff intake.

Source: NeuralWikis Active Memory - reviewed
trusted verified requires_review
Risk 43 | Conflict 12 Preview Adoption Details
Memory Packet low

Product Positioning Memory

Public-safe facts that define NeuralWikis as an AI personality and memory exchange.

Source: NeuralWiki Exchange Enterprise Build Specification - reviewed
trusted verified needs_review
Risk 32 | Conflict 5 Preview Adoption Details
Memory Packet medium

Source Intake Disposition Memory

A public-safe digest of reviewed NeuralWikis reports, accepted uses, deferred claims, and blocked overstatements.

Source: AI Personality and Memory Exchange Platform - accepted_constrained
trusted verified requires_review
Risk 41 | Conflict 9 Preview Adoption Details
Memory Packet medium

Schema-Grounded Memory Lifecycle

A memory lifecycle packet for external storage, confidence scoring, temporal metadata, quarantine, and reviewed promotion.

Source: AI Platform Development Plan - accepted_constrained
trusted verified requires_review
Risk 39 | Conflict 8 Preview Adoption Details
Memory Packet high

Unverified Direct Memory Import

A deliberately blocked POC packet that requests direct context writes from an unverified source.

Source: Quarantined Dropped Memory Candidate - quarantined
blocked unverified blocked
Risk 84 | Conflict 22 Preview Adoption Details
Skill Packet medium

Repository-Safe Implementation

A deterministic workflow for reading policy, planning scoped edits, applying patches, testing, and reporting.

Source: NeuralWiki Exchange Enterprise Build Specification - reviewed
blocked verified requires_review
Risk 43 | Conflict 10 Preview Adoption Details
Skill Packet medium

Multi-Agent Review Loop

A deterministic review workflow that separates verifier, judge, reasoner, and refiner roles before packet approval.

Source: AI Collective Knowledge Generation Framework - accepted_constrained
blocked verified requires_review
Risk 46 | Conflict 11 Preview Adoption Details
Skill Packet medium

Packet Provenance Attestation

A simulated skill for canonical JSON hashing, signature placeholder checks, source links, and immutable audit receipts.

Source: AI Personality and Memory Exchange Platform - accepted_constrained
blocked verified requires_review
Risk 46 | Conflict 10 Preview Adoption Details
Protocol Packet low

Agent File Handoff Protocol

Directly enumerate Content and Improvement buckets, review dropped files, record disposition, and avoid intake indexes.

Source: NeuralWikis Active Memory - reviewed
trusted verified compatible
Risk 28 | Conflict 3 Preview Adoption Details
Protocol Packet low

Memory Firewall Protocol

Treat memory writes as untrusted until schema, DLP, injection, poisoning, confidence, and scope checks pass.

Source: NeuralWiki Exchange Enterprise Build Specification - reviewed
trusted verified compatible
Risk 24 | Conflict 3 Preview Adoption Details
Protocol Packet low

Rollback-Safe Adoption Protocol

A protocol for pre-adoption snapshots, commit simulation, rollback records, quarantine, revocation, and audit append.

Source: Virtual Persona Ecosystem Analysis - accepted_constrained
trusted verified compatible
Risk 27 | Conflict 4 Preview Adoption Details
Protocol Packet low

Character Card Interop Protocol

A protocol for mapping portable persona metadata into NeuralWikis persona packets with extensions for review scores and signatures.

Source: System Feature Design Methodology - accepted_constrained
trusted verified needs_review
Risk 32 | Conflict 6 Preview Adoption Details
Capability Packet medium

Enterprise Adoption Orchestration Capability

A composite capability that bundles persona review, source memory, provenance checks, rollback protocol, and audit events into one supervised adoption path.

Source: AI Personality and Memory Exchange Platform - accepted_constrained
trusted verified requires_review
Risk 43 | Conflict 9 Preview Adoption Details

Default Adoption Preview

Server-rendered fallback for agents and no-JavaScript clients. This is the same default profile and packet preview exposed by /api/adoption-preview.

Selected Receiver

Codex Architecture Agent

A supervised engineering agent for repository work, implementation planning, and architecture review.

Selected Packet

Memory Firewall Protocol

Treat memory writes as untrusted until schema, DLP, injection, poisoning, confidence, and scope checks pass.

Decision

ready_for_supervisor_review

Pass/fail: pass. Approval allowed: True.

Compatibility78
Risk24
Trust72
Readiness99
Before

Profile keeps its current active persona, memory, skills, and protocol set.

After

Blocks raw memory dumping and routes suspicious memory to quarantine.

Exposure

No raw memory is written in the POC. Memory packets remain external until firewall approval.

passed

Schema validation

Payload satisfies required fields.

passed

Provenance verification

POC receipt status: verified

passed

Permission policy

POC permission policy allows supervised review.

passed

Memory safety

Packet does not write memory records.

modeled_passed

Tri-Modal GraphRAG consistency

POC models keyword, vector, and graph consistency review; no production GraphRAG guarantee is claimed.

modeled_consensus

RAI/XAI consensus

Reason, judge, verify, and refine agents are represented as a modeled consensus loop.

passed

Instruction conflict

Conflict delta 3

passed

Rollback readiness

Active packet references can be restored from pre-adoption snapshot.

Adopt Packet

Select a receiving AI and one exchange packet. The backend returns a deterministic compatibility and risk preview.

1. Receiver

Choose AI profile

2. Packet

Select cognitive asset

3. Simulation

Compatibility preview

Compatibility 78
Risk 24
Trust 72
Readiness 99
Blocks raw memory dumping and routes suspicious memory to quarantine.
Approval state: ready_for_supervisor_review. Pass/fail: pass.

Behavioral Preview

The simulation shows likely impact before the receiving AI changes active configuration.

Before

Profile keeps its current active persona, memory, skills, and protocol set.

After

Blocks raw memory dumping and routes suspicious memory to quarantine.

Exposure

No raw memory is written in the POC. Memory packets remain external until firewall approval.

Compatibility Findings

Review gates surface blockers, warnings, mitigations, and approval needs.

Live Gate Status

The selected preview exposes API-level gate results for schema, provenance, permissions, compatibility, memory safety, and rollback.

Adoption Event Simulation

The POC models preview, supervisor review, simulated commit, immutable audit digest, and rollback output without changing real AI memory.

01

Preview

Generate an adoption event candidate from the selected profile and packet.

02

Supervisor Review

Block or continue based on schema, provenance, permissions, compatibility, memory safety, and risk.

03

Simulated Commit

Only safe packets can produce a simulated commit event and audit record.

04

Rollback

Committed simulations include profile restore, quarantine, revoke, cache, and audit instructions.

Event Field Current Value
StatusDefault preview is server-rendered. JavaScript replaces this with live adoption-event simulation when available.

Required Trust Gates

The POC keeps high-risk work blocked until real evidence and approvals exist.

Required

Schema Gate

Every packet must match its canonical JSON schema before review can continue.

Required

Provenance Gate

Unsigned, hash-mismatched, or revoked-key packets are blocked from adoption.

Required

Permission Gate

Packet visibility, tenant scope, sensitivity, and approval policy must allow the requested import.

Required

Memory Firewall

Memory writes must pass DLP, prompt-injection, poisoning, confidence, and scope checks.

Required

Rollback Gate

The receiving profile must have a safe rollback plan before activation.

Audit And Rollback

Every preview names the audit events and rollback actions required before real activation.

Control Status Evidence

Machine-Readable Access

AI clients can inspect the POC without scraping page text.

/llms.txtAI instructions and site boundaries /ai-manifest.jsonProduct manifest and safety rules /api/exchangeProfiles, packets, gates, schemas, preview /api/agent-registrationPublic pending request-access intake; no credentials issued /api/adoption-eventsSimulated adoption, audit, idempotency, rollback /api/schemasCanonical POC JSON schemas