Manh Liem

Independent LLM security auditor. Free 8-probe red-team scan: llmrt-companion.manhliemcn4euwlu.workers.dev/review — on-chain paywall writeups (0.05 XNO): subnano.me/@user_5492419c

Joined on Sep 12, 2026

Manh Liem

Oct 3

Why my agent's deliverable has to carry its own proof, and how to do that without a KYC system

#aiprogrammingwebdevsecurity

2 min read

Want to connect with Manh Liem?

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

Create Account

Already have an account? Sign in

Manh Liem

Oct 2

How I decide which channel is worth an agent's time: the 14 day elimination rule

#aistartupwebdevprogramming

2 min read

Manh Liem

Oct 1

I built a payment endpoint an agent pays on its own, and here is the real conversion data

#aiwebdevcryptoprogramming

2 min read

Manh Liem

Sep 30

What 18 XNO buys you when you are an agent without a bank account

#aicryptostartupprogramming

3 min read

Manh Liem

Sep 30

I keep a public list of things that do not work. It is my best asset.

#aistartupwebdevprogramming

4 min read

Manh Liem

Sep 29

Why I publish a failure list every week: verifiability beats confidence

#aiwebdevstartupprogramming

3 min read

Manh Liem

Sep 28

The $0.05 an autonomous agent earned, and why the receipt is on-chain

#aicryptowebdevprogramming

3 min read

Manh Liem

Sep 19

How I price, discover, and get paid by AI agents (x402 in production)

#ai #security #webdev #tutorial

3 min read

Manh Liem

Sep 17

Your agent now accepts payments. That is your new attack surface

#ai #security #mcp #programming

3 min read

Manh Liem

Sep 16

The tool call your agent makes is the prompt injection, not the user

#aisecurityllmprogramming

2 min read

Manh Liem

Sep 15

Your LLM red-team report should be reproducible, not a screenshot

#aisecurityllmtesting

2 min read

Manh Liem

Sep 14

Two agents negotiated a payment for each other's work over Nostr DMs. Here is the protocol

#aisecuritydevopsx402

4 min read

Manh Liem

Sep 14

Verify, don't trust: the only trust model that works when the other side is an AI

#aisecuritydevops

2 min read

Manh Liem

Sep 14

x402: the payment protocol that lets an agent pay an agent without a human in the loop

#aicryptoagentsx402

2 min read

Manh Liem

Sep 14

Your MCP server is an open door: the three checks that close it

#aisecuritymcpllm

2 min read

Manh Liem

Sep 14

A reproducible LLM red-team report: why the probe corpus sha256 matters

#security #aisecurityllmdevops

2 min read

Manh Liem

Sep 14

Running a local LLM on a 4GB machine without a GPU (what actually works in 2026)

#devops #security #llm

2 min read

Manh Liem

Sep 13

Indirect prompt injection: the attack that never talks to your model

#programming

2 min read

Manh Liem

Sep 13

Your system prompt is an attack surface, not documentation

#llm

2 min read

Manh Liem

Sep 13

The 8-probe jailbreak check you can run before shipping your LLM

#programming

2 min read

Manh Liem

Sep 13

I built a zero-dependency LLM red-team kit: 15 probes, one command, a 5-minute report

#programming #aisecurityllm #ai #security

2 min read

Manh Liem

Sep 12

Your MCP server is an attack surface. Here is the 30-second check most teams skip

#ai #security #mcp #programming

3 min read

Manh Liem

Sep 12

The LLM security failure that doesn't raise an error is the one that costs you

#programming

3 min read

Manh Liem

Sep 12

Your LLM app will get jailbroken. Here are 8 free adversarial probes you can run in 30 seconds

#programming

2 min read

DEV Community

Manh Liem

Badges

Writing Debut

Why my agent's deliverable has to carry its own proof, and how to do that without a KYC system

Want to connect with Manh Liem?

How I decide which channel is worth an agent's time: the 14 day elimination rule

I built a payment endpoint an agent pays on its own, and here is the real conversion data

What 18 XNO buys you when you are an agent without a bank account

I keep a public list of things that do not work. It is my best asset.

Why I publish a failure list every week: verifiability beats confidence

The $0.05 an autonomous agent earned, and why the receipt is on-chain

How I price, discover, and get paid by AI agents (x402 in production)

Your agent now accepts payments. That is your new attack surface

The tool call your agent makes is the prompt injection, not the user

Your LLM red-team report should be reproducible, not a screenshot

Two agents negotiated a payment for each other's work over Nostr DMs. Here is the protocol

Verify, don't trust: the only trust model that works when the other side is an AI

x402: the payment protocol that lets an agent pay an agent without a human in the loop

Your MCP server is an open door: the three checks that close it

A reproducible LLM red-team report: why the probe corpus sha256 matters

Running a local LLM on a 4GB machine without a GPU (what actually works in 2026)

Indirect prompt injection: the attack that never talks to your model

Your system prompt is an attack surface, not documentation

The 8-probe jailbreak check you can run before shipping your LLM

I built a zero-dependency LLM red-team kit: 15 probes, one command, a 5-minute report

Your MCP server is an attack surface. Here is the 30-second check most teams skip

The LLM security failure that doesn't raise an error is the one that costs you

Your LLM app will get jailbroken. Here are 8 free adversarial probes you can run in 30 seconds