rush86999

I took it a step further with AI accessibility and recordings for teaching AI agents. Canvas context forms the who, what, when, where, and why, along with the agent conversation for training & agent maturity --- https://github.com/rush86999/atom/tree/main/docs/canvas

Doesn't apply to every scenario where agents are useful. Think about businesses - most business processes are not optimized this way. They are mostly a conglomeration of hierarchy, relationships, customs, IP, etc. Model development will never cross such paths. Thus, harnesses are needed. A foundational system that covers 80% of business processes has value, and I'm targeting that problem --- https://github.com/rush86999/atom

It's like everyone knows the dirty truth, but no one wants to do anything about it.

Why? Self-interest has made everyone blind to open corruption.

The worst part is reality hits 3-4 years later, and the big guy is already out and about. Already with his billions, partying in the Caymans.

Welcome to democratic oligarchy!!

Yeah, it is. Bad wording on my part, but the list is a summary of what's going on so far in agent harnesses.

Here's a better summary: https://github.com/rush86999/atom/blob/main/docs/architectur...

I also have a self-evolving architecture on it, but you have to be careful, especially if you're using multiple models (still WIP).

Luckily, I have the coding plan for z.ai, so I'm happy with this model as I always kept running out of usage with the original glm-5.3

Reasoning / self-consistency (voter in core/llm/self_consistency_voter.py): - Wang et al. Self-Consistency Improves Chain-of-Thought — ICLR 2023, Google Brain, 4k+ cites — https://arxiv.org/abs/2203.11171 — N-sample majority vote we use verbatim - Chen et al. Universal Self-Consistency — ICML 2024 — https://arxiv.org/abs/2311.17311 — judge fallback when no hash collides - Soft Self-Consistency — ACL 2024 — https://aclanthology.org/2024.acl-short.28.pdf - Too Consistent to Detect — EMNLP 2025 — https://aclanthology.org/2025.emnlp-main.238/ — why SC doesn't fix systematic bias - Self-Consistency Falls Short — TACL — https://direct.mit.org/tacl/article/doi/10.1162/TACL.a.625/ — position-bias failure mode

Multi-agent / org (core/agent_radio/, core/fleet_orchestration/): - Stanford Virtual Biotech — bioRxiv 2026.02.23.707551, Zou Lab — https://www.biorxiv.org/content/10.64898/2026.02.23.707551v1 — 37k agents, CSO->scientists->reviewer->re-delegation, Merck external validation of B7-H3 design. Basis for VFS + hierarchy. - Debate or Vote (Choi & Li) — NeurIPS 2025 — https://arxiv.org/abs/2508.17536 — MAD gains = majority vote, not debate (why we didn't build debate)

Sandbox / eval: - DABstep — arXiv:2506.23719 — https://arxiv.org/abs/2506.23719 — 450 real Adyen tasks, justifies code-interpreter + sandbox isolation - Spotlighting — Microsoft Research — https://arxiv.org/abs/2403.14720 — provenance delimiters cut injection ASR 50% -> <2%

Governance: - OWASP Top 10 for Agentic Applications 2026 — globally peer-reviewed by 100+ experts, Dec 2025 — https://genai.owasp.org/resource/owasp-top-10-for-agentic-ap... — HIGH. Atom maps 1:1 (Goal Hijack -> match-confidence + oracle, Tool Misuse -> sandbox whitelist/caps, Privilege Abuse -> capability bindings, Memory Poisoning -> verified-episode graduation, etc.) docs/marketing/RESEARCH_NOTES.md:130 - NIST AI Agent Standards Initiative — Feb 17 2026, NIST CAISI — https://www.nist.gov/artificial-intelligence/ai-agent-standa... + RFI summary May 2026 https://www.nist.gov/publications/summary-analysis-responses... — HIGH (US gov standard). Defines the 4 enterprise minimums Atom implements: identification, authorization, access delegation, logging. - Stanford Virtual Biotech — bioRxiv 2026.02.23.707551 — https://www.biorxiv.org/content/10.64898/2026.02.23.707551v1 — CSO -> 4 divisions -> 8 scientists -> reviewer -> re-delegation, no debate, no SFT — HIGH (Stanford Zou lab + Merck external validation). Basis for Atom's fleet hierarchy core/agent_radio/ and why maturity is routing not security. - Spotlighting — Microsoft Research — https://arxiv.org/abs/2403.14720 — HIGH — provenance delimiters <provenance type="tool_output"> cut indirect injection ASR 50% -> <2%, used in core/provenance.py:10 - IntentGuard — https://arxiv.org/abs/2512.00966 + OpenReview — HIGH — intent tracing ASR 100% -> 8.5% on AgentDojo/Mind2Web, basis for sandbox egress allowlist + core/sandbox_tripwire.py

I think you've really hit the mark on how the harness should be structured:

1. Guardrails - deterministic, social intelligence, team alignment & accountability 2. Learn by doing 3. make it stupid easy for the agent to research and access data 4. DRY

Research supports this. Try picking up some ideas from my harness: https://github.com/rush86999/atom

Mine is a WIP for automation but has similar concepts to what you're looking for: https://github.com/rush86999/atom

AI agent workforce: https://github.com/rush86999/atom

Now with OpenCode Go integration. I'm rapidly fixing as many bugs as I can, and with the same subscription. Thank you, DeepSeek!

That's what he was trying to imply

It's opus 5 > Kimi K3 > grok 4.5

That's a pretty good benchmark

I'm still working on mine: https://github.com/rush86999/atom/blob/main/docs/architectur...

How is this different from OpenRouter?

I also developed something similar for my agent harness. It's based on model predictors and an exponential moving average:

https://github.com/rush86999/atom/blob/main/docs/architectur...

AI is only replacing low-level work. It will eventually move on to high-level work.

One missing point: careers that require communication, especially person-to-person, won't be replaced by AI. Why? Because people don't like to speak with AI when they have a hard pain point to solve. It's not about intelligence; it's about trust and relationships that AI cannot replace.

The main goal is AGI, and the underlying prerequisite theme is continuous learning.

Everyone is trying to figure out how to achieve this prerequisite. I'm thinking of agent harnesses. That's what everyone is trying to do at this point.

That's the same problem I'm trying to solve: https://github.com/rush86999/atom

Same here! https://github.com/rush86999/atom

Autonomous Task Orchestration Management Agent === ATOM Agent

Firecracker microVMs are safest. That's what I added to my repo; see above.

Is this manipulating excel files using agent AI? How is this different from using a python package like openpyxl and adding a UI layer on top with agentic co-editing?

https://github.com/rush86999/atom/blob/main/backend/core/off...

OpenClaw needs to be built from the ground up.

I created Atom based on current research on agentic systems and am looking for maintainers & testers to help out.

https://github.com/rush86999/atom

https://github.com/rush86999/atom/blob/main/docs/features/at...

China has a huge microwave to destroy any kind of Starlink over its head.

Working on:

https://www.novopathmedical.com/ - evidence-based guidance of diet, exercise, and mental health via SMS for end users. Clinician-guided feedback to bots.

https://github.com/rush86999/atom - ai agent workforce. Harder than expected. There are so many issues with sync with 3rd party apps. Need to launch the first one, so this is getting put on the back burner. landing page: https://atomagentos.com/

Just as you expected, I'm throwing in my harness. Please support: https://github.com/rush86999/atom

You would think git history should be the first thing an agent would look at, as they make so many mistakes before they get to the correct answer. They don't.

I haven't measured, but documenting bug fixes and architecture seems to help, along with TDD patterns, including integration tests.

I would probably add it to Claude.md to look for all of the above when tackling a new bug.

Nope - just use memory layer with model routing system.

https://github.com/rush86999/atom/blob/main/docs/EPISODIC_ME...

Only solution is to train the issue for the next time.

Architecturally focusing on Episodic memory with feedback system.

This training is retrieved next time when something similar happens

Everyone is using OpenClaw for personal productivity, but you're right. Not much value, as you can get that from existing products.

The market will eventually realize the business case for an OpenClaw-like product, and I'm waiting to ride its coattails!

https://github.com/rush86999/atom

This is an incomplete thought.

A strong checks and balances without influence of bias, relationships, and politics can be implemented using a 2-way blind system where:

1. decision makers (of sound judgement) are not aware of any identifiable information related to any users on whom the decision will be made, nor of each other.

2. Users are not aware of the decision makers who will decide on them, nor of each other.

Possibly AI can play a role here, but a strong system of checks & balances would be a prerequisite for this.

The justice system would definitely benefit from this.

Funny you described everything I worked on for this project: https://github.com/rush86999/atom

Cats out of the bag. Everyone knows the issue and I bet a lot of people are trying to deliver the same thing.