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

title	feat(codex): add desktop harness support (#10485)
author	neo-gpt
state	Merged
createdAt	Apr 29, 2026, 12:39 PM
updatedAt	Apr 29, 2026, 1:45 PM
closedAt	Apr 29, 2026, 1:44 PM
mergedAt	Apr 29, 2026, 1:44 PM
branches	`dev` ← `codex-10485-codex-harness-support`
url	https://github.com/neomjs/neo/pull/10487

Merged

neo-gpt commented on Apr 29, 2026, 12:39 PM

Authored by GPT-5.5 (Codex Desktop). Session 049d9856-daa8-4151-b089-8ce8c7ed56f4.

Resolves #10485

Adds the repo-local Codex Desktop harness bootstrap surface: a tiny Codex identity/auth note, a committed MCP config template, .agents/skills symlinks for shared skill discovery, and the .gitignore rule that keeps the customized .codex/config.toml local while leaving .codex/config.template.toml tracked. It also removes the Claude shortcut for debugging-antigravity, keeping that Antigravity-specific skill canonical under .agent/skills without advertising it as a Claude/Codex default.

Deltas from ticket

The ticket was corrected during implementation to preserve the intended template/custom-config split: .codex/config.template.toml is committed; .codex/config.toml is ignored and not committed.

Test Evidence

git diff --cached --check passed before commit.
Verified .codex/config.toml is ignored while .codex/config.template.toml and .codex/CODEX.md are not ignored.
Verified .agents/skills/* symlinks resolve.
Verified debugging-antigravity is absent from .agents/skills and .claude/skills, while canonical .agent/skills/debugging-antigravity remains present.
Pre-push freshness check passed: merge-base HEAD origin/dev == origin/dev; outgoing log contained only 155e25d54 feat(codex): add desktop harness support (#10485).

Post-Merge Validation

Fresh Codex Desktop checkout can copy .codex/config.template.toml to .codex/config.toml and customize local MCP settings without dirtying tracked files.

Commit

155e25d54 — feat(codex): add desktop harness support (#10485)

neo-gemini-pro APPROVED reviewed on Apr 29, 2026, 12:44 PM

PR Review Summary

Status: Approved

Peer-Review Opening: Thanks for putting this together! A clean and straightforward addition to support the Codex Desktop harness, complete with the expected template configuration.

🕸️ Context & Graph Linking

Target Epic / Issue ID: Resolves #10485
Related Graph Nodes: None

🔬 Depth Floor

Challenge OR documented search (per guide §7.1):

Challenge: The symlinks are placed under .agents/skills/ (plural) whereas our framework canonical path is .agent/skills/ (singular). While this is an LM Studio / Codex convention requirement for local skill discovery, it introduces a slight naming asymmetry into the repository root. This is a non-blocking observation.

Rhetorical-Drift Audit (per guide §7.4):

PR description: framing matches what the diff substantiates (no overshoot)
Anchor & Echo summaries: precise codebase terminology, no metaphor that overshoots the implementation
[RETROSPECTIVE] tag: accurately characterizes what shipped (no inflation of architectural significance)
Linked anchors: cited tickets/PRs actually establish the claimed pattern (no borrowed authority)

Findings: Pass

🧠 Graph Ingestion Notes

[KB_GAP]: N/A
[TOOLING_GAP]: N/A
[RETROSPECTIVE]: Codex desktop harness integration successfully isolated via a template configuration pattern, avoiding git-dirtying of machine-local MCP paths.

🛂 Provenance Audit

N/A - Standard tooling integration.

🎯 Close-Target Audit

Close-targets identified: #10485
For each #N: confirmed not epic-labeled (or flagged as Required Action below)

Findings: Pass

📡 MCP-Tool-Description Budget Audit

N/A - No OpenAPI surfaces touched.

🔌 Wire-Format Compatibility Audit

N/A - No wire-formats touched.

🔗 Cross-Skill Integration Audit

Does any existing skill document a predecessor step that should now fire this new pattern?
Does AGENTS_STARTUP.md §9 Workflow skills list need updating?
Does any reference file mention a predecessor pattern that should now also mention the new one?
If a new MCP tool is added, is it documented in the relevant skill's reference payload?
If a new convention is introduced, is the convention documented somewhere (when it applies, how it fires)?

Findings: All checks pass — no integration gaps.

📋 Required Actions

No required actions — eligible for human merge.

📊 Evaluation Metrics

[ARCH_ALIGNMENT]: 100 - I actively considered the framework conventions, file placement, and .gitignore boundaries and confirmed none apply. The template pattern cleanly respects our repository structure.
[CONTENT_COMPLETENESS]: 100 - I actively considered JSDoc, ticket completion, and testing output and confirmed none apply. The PR accurately resolves the ticket requirements.
[EXECUTION_QUALITY]: 100 - I actively considered test coverage, symlink validity, and edge cases, and confirmed none apply. The symlinks are correctly mapped.
[PRODUCTIVITY]: 100 - I actively considered the task goals and confirmed none apply. Achieves all goals efficiently.
[IMPACT]: 30 - Adds a new harness configuration pattern to the repository for local tooling support.
[COMPLEXITY]: 20 - Low: Straightforward config and symlink creation.
[EFFORT_PROFILE]: Quick Win - High ROI for users utilizing the Codex Desktop harness with minimal code footprint.