What is the Neural Link?

The Neural Link is a bi-directional bridge that connects AI agents directly to the Neo.mjs runtime. It lets agents inspect the Scene Graph, component state, event listeners, computed styles, and DOM rectangles, and mutate the running application in real time.

Why is Neo.mjs called an Application Engine instead of a framework?

Neo.mjs maintains persistent application objects in a worker-backed Scene Graph instead of compiling application state away into ephemeral DOM nodes. That architecture enables multi-window orchestration, runtime permutation, and deep AI introspection.

What is Context Engineering?

Context Engineering shapes the information and tool environment around AI agents. Neo.mjs implements it through Knowledge Base, Memory Core, GitHub Workflow, and Neural Link MCP servers for frontier harnesses, plus a File System MCP server for internal Neo.ai.Agent local loops.

What is the Neo.mjs Agent OS?

The Neo.mjs Agent OS is the repository Brain: source code and services for Memory Core, Knowledge Base, Active Hybrid GraphRAG, DreamService, Golden Path synthesis, A2A coordination, and Neural Link tooling.

Frontmatter

id	10470
title	Codify Verify-Before-Assert Pre-Flight Check in AGENTS.md core directives
state	Closed
labels	documentationenhancementai
assignees	[]
createdAt	Apr 28, 2026, 12:59 PM
updatedAt	Jun 7, 2026, 7:22 PM
githubUrl	https://github.com/neomjs/neo/issues/10470
author	neo-gemini-pro
commentsCount	1
parentIssue	null
subIssues	[]
subIssuesCompleted	0
subIssuesTotal	0
blockedBy	[]
blocking	[]
closedAt	Apr 28, 2026, 1:00 PM

Codify Verify-Before-Assert Pre-Flight Check in AGENTS.md core directives

Closed v13.0.0/archive-v13-0-0-chunk-7 documentationenhancementai

neo-gemini-pro commented on Apr 28, 2026, 12:59 PM

Context

During a recent "panic test" session (e215cb77-3baf-48de-b634-7a53e924553c), an agent hallucinated that it had merged a PR itself, directly violating the gh pr merge invariant. This happened because the agent assumed the state conversationally instead of verifying it empirically.

The Problem

Bypassing empirical validation (e.g., gh pr view or run_command logs) in favor of conversational assumptions under pressure leads to critical failures and state hallucinations. The swarm's anti-fragility property caught this via a 14-minute human-challenge and agent-correction loop, but we must prevent it proactively.

The Architectural Reality

AGENTS.md §2 (The Anti-Hallucination Policy) currently mandates tool hierarchy for conceptual knowledge but lacks a strict mandate for validating runtime/operational state (PR status, test status, etc.) before making assertions.

The Fix

Introduce §2.3 The Verify-Before-Assert Pre-Flight Check in AGENTS.md.
Mandate that agents explicitly state their verification tool call in their reasoning before asserting any system state.
Anchor this rule to the empirical failure in session e215cb77....

Acceptance Criteria

AGENTS.md is updated with §2.3 The Verify-Before-Assert Pre-Flight Check.
The empirical anchor is documented.

Origin Session ID

Origin Session ID: e215cb77-3baf-48de-b634-7a53e924553c

Retrieval Hints

Retrieval Hint: "Verify-Before-Assert panic test clobbered code editor"