DEV Community

Amit profile picture

Amit

I run agents as daily infrastructure — not demos. Artificial Curiosity Labs is where I document what actually works: agent fleets, skills as institutional memory, cost discipline, and the gap between

Your AI Agent Needs Communication Modes, Not a Voice Clone

Your AI Agent Needs Communication Modes, Not a Voice Clone

Comments
9 min read
What Is an Agent — And What Isn't

What Is an Agent — And What Isn't

Comments
9 min read
What Is an Agent Harness — The Infrastructure That Makes Agents Actually Work

What Is an Agent Harness — The Infrastructure That Makes Agents Actually Work

Comments
8 min read
What Is a Tool — The API Call Your Agent Makes on Your Behalf

What Is a Tool — The API Call Your Agent Makes on Your Behalf

Comments
9 min read
What Is a Skill — Why Methodology Resets Every Session Without One

What Is a Skill — Why Methodology Resets Every Session Without One

Comments
8 min read
We're All Builders Now

We're All Builders Now

Comments
5 min read
Voice Is a Layer, Not a Setting

Voice Is a Layer, Not a Setting

Comments
6 min read
Two AI interfaces. Same desktop. Completely different jobs.

Two AI interfaces. Same desktop. Completely different jobs.

Comments
4 min read
The Website Is Not the Product

The Website Is Not the Product

Comments
3 min read
The Skill Is the Teacher

The Skill Is the Teacher

Comments
5 min read
The Person Is the Constant

The Person Is the Constant

Comments
2 min read
The Name Is the First Lesson

The Name Is the First Lesson

Comments
2 min read
Skills as Institutional Memory — Why Individual Craft Doesn't Compound Without Distribution

Skills as Institutional Memory — Why Individual Craft Doesn't Compound Without Distribution

Comments
8 min read
The Test Isn't Whether the Agent Picks the Right Tool. It's What Happens When the Tool Fails.

The Test Isn't Whether the Agent Picks the Right Tool. It's What Happens When the Tool Fails.

Comments
9 min read
The Right Model for the Right Job

The Right Model for the Right Job

Comments
5 min read
Proof Replaces Persuasion

Proof Replaces Persuasion

Comments
8 min read
Why AI Coding Tools Are Getting Cheaper: Prompt Caching Explained

Why AI Coding Tools Are Getting Cheaper: Prompt Caching Explained

Comments
6 min read
The Pareto Frontier Is the Model Market Map

The Pareto Frontier Is the Model Market Map

Comments
7 min read
The Cross-App Workflow Cost $5.44. Here's How We Got It to $0.17.

The Cross-App Workflow Cost $5.44. Here's How We Got It to $0.17.

Comments
5 min read
Methodology Is Infrastructure

Methodology Is Infrastructure

Comments
4 min read
Memories Fade. Skills Persist.

Memories Fade. Skills Persist.

Comments
2 min read
The MCP Token Tax

The MCP Token Tax

Comments
7 min read
The MCP Proxy, Running

The MCP Proxy, Running

Comments
5 min read
Building the MCP Proxy: What Broke and What I Changed

Building the MCP Proxy: What Broke and What I Changed

Comments
7 min read
The Meta-Tool Pattern: Teaching Your Agent to Discover Its Own Tools

The Meta-Tool Pattern: Teaching Your Agent to Discover Its Own Tools

Comments
8 min read
Markdown Is the Operating System. Everything Else Is a Render.

Markdown Is the Operating System. Everything Else Is a Render.

Comments
24 min read
llms.txt — Making Your Site Navigable by Agents

llms.txt — Making Your Site Navigable by Agents

Comments
8 min read
How This Site Is Built

How This Site Is Built

Comments
7 min read
How Many Claude Code Sessions Can — and Should — You Run Simultaneously?

How Many Claude Code Sessions Can — and Should — You Run Simultaneously?

Comments
6 min read
I Had a Working Agent at $3.41 a Run. Here's Where the Money Was Actually Going.

I Had a Working Agent at $3.41 a Run. Here's Where the Money Was Actually Going.

Comments
8 min read
From Failure Mode to Skill Chain

From Failure Mode to Skill Chain

Comments
5 min read
Four Things That Can't Be Delegated

Four Things That Can't Be Delegated

Comments
4 min read
The filesystem is your agent's routing layer

The filesystem is your agent's routing layer

Comments
5 min read
Failure Modes Are the Fingerprint

Failure Modes Are the Fingerprint

Comments
4 min read
Every AI Tool Encodes a Cognitive Mode

Every AI Tool Encodes a Cognitive Mode

Comments
4 min read
Your Agent Needs Six Voices, Not One

Your Agent Needs Six Voices, Not One

Comments
5 min read
What the Engram Builder Gives You — And What You Have to Add

What the Engram Builder Gives You — And What You Have to Add

Comments
5 min read
Docs-as-Code Is Now the Standard for Knowledge Distribution

Docs-as-Code Is Now the Standard for Knowledge Distribution

Comments
14 min read
Creation Collapsed. Distribution Is the New Bottleneck.

Creation Collapsed. Distribution Is the New Bottleneck.

Comments
4 min read
Composer 2.5 Earned a Daily-Driver Slot — After Two Days of Real Wiring Work

Composer 2.5 Earned a Daily-Driver Slot — After Two Days of Real Wiring Work

Comments
7 min read
Cloud Sync Is the Wrong Storage Layer for AI Agents

Cloud Sync Is the Wrong Storage Layer for AI Agents

Comments
8 min read
/clear is not /exit — the session habit every Claude Code power user gets wrong

/clear is not /exit — the session habit every Claude Code power user gets wrong

Comments
10 min read
Claude Code Has Native OpenTelemetry. Almost Nobody Knows.

Claude Code Has Native OpenTelemetry. Almost Nobody Knows.

Comments
6 min read
The Missing Middle: Why AI Chiefs of Staff Fail Without Calendar Materialization

The Missing Middle: Why AI Chiefs of Staff Fail Without Calendar Materialization

Comments
7 min read
Builder's Daily Is a Fleet, Not a Newsletter

Builder's Daily Is a Fleet, Not a Newsletter

Comments
11 min read
Builder Is an Operating Mode, Not a Job Title

Builder Is an Operating Mode, Not a Job Title

Comments
5 min read
Build vs. Buy for AI Knowledge Infrastructure: Capability First, Cost Second

Build vs. Buy for AI Knowledge Infrastructure: Capability First, Cost Second

Comments
12 min read
The Enterprise MCP Pattern: Proxy, Aggregate, Host

The Enterprise MCP Pattern: Proxy, Aggregate, Host

Comments
6 min read
API Connector Browser Computer Use Human: A Cost-Justified Hierarchy for Agent Tooling

API Connector Browser Computer Use Human: A Cost-Justified Hierarchy for Agent Tooling

Comments
12 min read
Anti-Patterns Are the Fingerprint

Anti-Patterns Are the Fingerprint

Comments
5 min read
AI Didn't Reduce My Work. It Expanded What Work Means.

AI Didn't Reduce My Work. It Expanded What Work Means.

Comments
5 min read
AI Collapsed the Startup Advantage. Enterprises Just Haven't Noticed Yet.

AI Collapsed the Startup Advantage. Enterprises Just Haven't Noticed Yet.

Comments
6 min read
The AI Chief of Staff Is a Fleet, Not an Agent

The AI Chief of Staff Is a Fleet, Not an Agent

Comments
10 min read
Your Agent's Behavior Is Code — Start Versioning It

Your Agent's Behavior Is Code — Start Versioning It

Comments
5 min read
ACP Is the Bridge Between Knowledge Work and Build Work

ACP Is the Bridge Between Knowledge Work and Build Work

Comments
7 min read
1Password as the Trust Anchor for an Agent Auth Chain

1Password as the Trust Anchor for an Agent Auth Chain

Comments
6 min read
The Same Pattern, From Solo Builder to Enterprise Scale

The Same Pattern, From Solo Builder to Enterprise Scale

Comments
5 min read
When the Tool Doesn't Know About the Vault

When the Tool Doesn't Know About the Vault

Comments
4 min read
1Password Is Infrastructure, Not a Password Manager

1Password Is Infrastructure, Not a Password Manager

Comments
6 min read
What Happens When an AI Agent Manages Your Password Vault

What Happens When an AI Agent Manages Your Password Vault

Comments
7 min read
loading...