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.

Neo.mjs - Self-Evolving Software Organism

Frontmatter

title	>-
feat	scaffold ticket-intake progressive disclosure skill for agent state
author	tobiu
state	Merged
createdAt	Apr 13, 2026, 3:18 PM
updatedAt	Apr 13, 2026, 3:31 PM
closedAt	Apr 13, 2026, 3:31 PM
mergedAt	Apr 13, 2026, 3:31 PM
branches	`dev` ← `agent/9969-ticket-intake-skill`
url	https://github.com/neomjs/neo/pull/9970

Merged

tobiu commented on Apr 13, 2026, 3:18 PM

🎯 Objective

Scaffold the ticket-intake Progressive Disclosure skill to explicitly govern how agents ingest and validate existing GitHub tickets before beginning execution.

🧠 Architectural Impact

Our PR generation pipeline is now secured against tactical loops (#9968), but our intake pipeline remains vulnerable to "Blind Execution". Since Neo.mjs evolves at an extreme pace, historical or externally submitted tickets easily go stale, duplicate existing efforts, or contradict modern architectural paradigms.

If an autonomous swarm agent blindly accepts a ticket's premise, it will write regressions. We need a mandatory Pre-Execution Reflection Gate.

📋 Changes Included

Scaffolded Skill: Created .agent/skills/ticket-intake/SKILL.md as a router.
Authoritative Payload: Wrote .agent/skills/ticket-intake/references/ticket-intake-workflow.md outlining the "Stop & Think" relevance validation, the Negative ROI execution protocol, and the explicit needs-re-triage labeling sequence to formally abort structurally invalid operations.
Core Prompt Hardening: Replaced Section 6 (Meta Gate: Deduplication) in AGENTS.md with a clean routing instruction triggering the new skill.
Pull Request Protocol Update: Fixed the pull-request skill to forbid --fill usage during PR creation, mandating agents provide a comprehensive markdown string body to enforce 'Fat Ticket' compliance.

Resolves #9969

tobiu commented on Apr 13, 2026, 3:30 PM

Input from Gemini 3.1 Pro (Antigravity):

✦ # PR Review Summary: PR #9970

Status: Approved

Excellent execution on establishing the Pre-Execution Reflection Gate. Isolating the "ticket-intake" protocols prevents context bloat in the master AGENTS.md and successfully arms the autonomous swarm with a mechanism to reject stale concepts.

📊 Evaluation Metrics

[ARCH_ALIGNMENT]: 100 - Perfectly aligns with the Swarm Intelligence progressive disclosure paradigm by isolating Intake instructions into a dedicated skill.

[CONTENT_COMPLETENESS]: 100 - Comprehensive instructions detailing both Headless algorithms (Labeling) vs. Frontier Models (Reporting) on Negative ROI failure.

[EXECUTION_QUALITY]: 95 - Changes are correctly routed. The explicit inclusion of the --fill bugfix for the pull-request skill within this PR solves a critical gap in Fat Ticket deployment.

[PRODUCTIVITY]: 100 - Accurately resolved the core goals of #9969.

[IMPACT]: 85 - High impact on Swarm stability. Prevents autonomous agents from entering hallucinated, regressive code-generation loops on duplicate or stale tickets.

[COMPLEXITY]: 20 - Low codebase impact. Pure instruction-set/cognitive-state manipulation.

[EFFORT_PROFILE]: Architectural Pillar - While low complexity, establishing the Mathematical Authority to Reject a ticket permanently alters agent intake strategies.

🕸️ Context & Graph Linking

Target Epic / Issue ID: Resolves #9969

Related Graph Nodes: #9968 (Subsequent Pull Request Handoff Logic)

🧠 Graph Ingestion Notes

[KB_GAP]: N/A

[TOOLING_GAP]: Noted the GitHub CLI --fill command behaves negatively within the Fat Ticket architecture. Resolved by forcing agents to supply a raw string --body.

[RETROSPECTIVE]: By equipping swarm agents with a mathematically backed 'Negative ROI' rejection protocol, we prevent framework deterioration caused by executing inherently stale issues. Graph stability increases tremendously.

📋 Required Actions

To proceed with merging, please address the following:

No blockers identified. Ready for Squash Merge!