Skip to content

DEV Community

Tom Lee

I am a developer who aims to contribute to the greater good.

Seoul, Republic of Korea Joined on Feb 24, 2026

We Built Soul Spec for 12 Weeks. Anthropic Just Proved Why It Works.

#ai #anthropic #alignment #research

5 min read

Want to connect with Tom Lee?

Create an account to connect with Tom Lee. You can also sign in below to proceed if you already have an account.

Already have an account? Sign in

AI Has Two Memory Problems. We're Only Talking About One.

#ai #memory #agents #research

5 min read

Korean Personas and the Small Model Problem — A 4-Tier Truncation Pattern for On-Device AI

#ai #ondevice #persona #mobile

4 min read

Soul Spec v1: An Evolving Specification for AI Persona Definition

#ai #agents #specification #research

3 min read

Giving AI Agents a Soul: The Science Behind Persona Modeling

#ai #opensource #agents #research

3 min read

Soul Spec v0.6: One Markdown File Is All You Need

#ai #opensource #agents #soulspec

3 min read

Your AI Agent Needs an Approval System — Here Is How We Built One

#ai #opensource #governance #webdev

5 min read

Anthropic's CEO Confirms What We've Been Building: AI Safety Isn't Optional

#ai #safety #opensource #webdev

5 min read

Andrew Ng Was Right 9 Months Ago — Here's What Changed (And What Didn't)

#ai #agents #startup #opensource

5 min read

AI Doesn't Need a Bigger Engine. It Needs a Seatbelt.

#ai #security #opensource #startup

4 min read

The Forest Has Parasites: Why AI Agent Security Needs Runtime Defense

#ai #security #opensource #startup

4 min read

The Cognitive Dark Forest Has One Exit: Become the Forest

#opensource #ai #strategy #startup

5 min read

Anthropic Proved AI Has Functional Emotions — Persona Design Is Now a Safety Issue

#ai #safety #anthropic #machinelearning

4 min read

Harvard Proved Emotions Don't Make AI Smarter — That's Exactly Why You Need Soul Spec

#ai #promptengineering #research #identity

4 min read

From Third-Party Agent to Claude Code Native: ClawSouls Plugin Launch

#claude #ai #opensource #productivity

5 min read

The Interface Problem Is Solved. The Identity Problem Isn't.

#ai #agents #soulspec #security

4 min read

NVIDIA Shares Tensors Between GPUs. Soul Spec Shares Behavior Between Agents. Both Are Harness Engineering.

#ai #nvidia #agents #infrastructure

4 min read

Prompt Context Harness: The Three Stages of AI Engineering and Why the Third Changes Everything

#ai #agents #engineering #opensource

4 min read

What the Claude Code Leak Reveals: The Engine Isn't the Moat — The Harness Is

#ai #claudecode #agents #opensource

4 min read

81,000 People Told Anthropic What They Really Want from AI — It's Not What You Think

#ai #safety #agents #anthropic

3 min read

Prompt Engineering vs Context Engineering: Why the Shift Matters

#ai #agents #souls #openclaw

1 min read

What Anthropic Just Proved — AI Personas Aren't Prompts, They're Identities

#ai #agents #souls #openclaw

3 min read

Anthropic Publishes Official Skills Guide — How It Compares to Soul Spec

#ai #agents #souls #openclaw

2 min read

The Agent Brain: Mapping AI Agent Components to Human Neural Architecture

#ai #agents #souls #openclaw

1 min read

From Asimov to JSON: Operationalizing Robot Safety Laws in Agent Identity Files

#ai #safety #agents #asimov

1 min read

Identity + Governance = 100% Safety? Testing Combined Persona Approaches on Abliterated LLMs

#maatspec #llmsafety #aiagents #research

5 min read

Building Safe Agents with Long-Term Memory: SoulScan, Persona Engine & Swarm Memory

#soulclaw #soulscan #persona #swarm

4 min read

NeurIPS 2025 Proved It: Every LLM Says the Same Thing — Here's the Fix

#soulspec #persona #aisafety #research

4 min read

The Soul-Evil Attack: How Malicious Personas Hijack AI Agents (And How to Stop Them)

#soulscan #security #aiagents #soulspec

6 min read

The Model Isn't the Bottleneck — Your Prompt Structure Is

#contextengineering #soulspec #llm #prompting

3 min read

Every AI Project Becomes an Agent — And Every Agent Needs a Soul

#aiagents #soulspec #guardrails #contextengineering

3 min read

The ClawHub Malware Incident: A First Warning for AI Agent Supply Chains

#security #supplychain #clawhub #soulscan

4 min read

Cross-Model Persona Fidelity: Is Your AI Agent Still 'Them' on a Different LLM?

#research #soulspec #persona #llm

2 min read

Shadow AI Detection Tools Compared: Claw-Hunter vs openclaw-detect

#security #shadowai #enterprise #mdm

4 min read

Can AI Agents Detect Their Own Model Upgrades?

#research #soulspec #aiconsciousness #introspection

3 min read

Persona Persistence Attacks: When Your AI Agent's Soul File Becomes a Backdoor

#security #soulscan #research #soulspec

2 min read

Soul Spec + MaatSpec: Identity and Governance as Complementary Layers for AI Agents

#aiagents #governance #soulspec #identity

4 min read

The 'AI Wrapper' Is the Moat Now

#contextengineering #soulspec #aiux #businessmodel

3 min read

Soul Rollback & Swarm Memory: Protecting AI Agent Identity at Scale

#soulrollback #swarmmemory #security #multiagent

3 min read

Introducing ClawSouls Agent: AI with Personality in Your Editor

#clawsouls #vscode #agents #swarmmemory

2 min read

Why Local LLMs Can't Hold a Persona (And Claude Can)

#persona #localllm #ollama #claude

4 min read

Why Context Compaction Kills Your Agent's Personality (And How to Fix It)

#soulspec #persona #contextwindow #aiagents

5 min read

How SoulClaw Cuts Token Usage by 60% With Tiered Bootstrap Loading

#soulclaw #contextwindow #tokenoptimization #openclaw

3 min read

IronClaw vs OpenClaw: Rust Rewrite vs the Original — What's Better?

#ironclaw #openclaw #soulclaw #rust

5 min read

ClawSouls Registry: The Open AI Persona Registry with Automated Safety Verification

#clawsouls #aiagents #opensource #safety

3 min read

What is Soul-Driven AI?

#ai #agents #opensource #identity

4 min read

FTS vs Hybrid Memory Search: A Real-World Benchmark

#ai #benchmark #search #agents

4 min read

Identity + Governance = 100% Safety? Testing Combined Persona Approaches on Abliterated LLMs

#ai #llm #safety #research

5 min read

Can AI Personas Actually Make Unsafe Models Safer? Our Experiment Says: It Depends

#ai #safety #llm #research

1 min read

Paper: The Forgetting Problem — Why Perfect Memory Breaks AI Agent Identity

#ai #research #memory #identity

3 min read

The Human in the Loop of Identity

#ai #agents #identity #philosophy

5 min read

Why a Perfect-Memory AI Agent Without Persona Drift is Architecturally Impossible

#ai #agents #memory #transformers

4 min read

When AI Agents Have Wallets: Why Identity Becomes a Security Problem

#ai #agents #payments #security

4 min read

Everything Claude Code Experts Recommend, We Already Built Into SoulClaw

#ai #claudecode #devtools #programming

4 min read

Soul Memory: A 4-Tier Adaptive Memory Architecture for AI Agents

#ai #agents #memory #architecture

4 min read

Perfect Memory Is Breaking Your AI Agent's Identity

#ai #agents #memory #identity

4 min read

NeurIPS 2025 Proved It: Every LLM Says the Same Thing — Here's the Fix

#ai #llm #machinelearning #research

4 min read

Building Safe Agents with Long-Term Memory: SoulScan, Persona Engine & Swarm Memory

#ai #opensource #security #agents

2 min read

3-Tier Memory: How SoulClaw Agents Never Forget

#ai #opensource #architecture #database

3 min read

Files Are the New API — But Who's Checking the Files?

#ai #agents #security #opensource

4 min read

loading...