Assili Salim

a vibe coder who is building something to reduce the API bill of using AI every months for companies , teams , developers and enterprises

Foussana , Kasserine , Tunisia Joined on Jun 9, 2026 https://github.com/salimassili62-afk/ai-costguard

Education

high school of Foussana , kasserine , Tunisia

Pronouns

Assili Salim

Aug 12

Multi-Agent Systems Share Context. They Should Share Budgets Too.

#ai #typescript #node #agents

2 min read

Want to connect with Assili Salim?

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

Create Account

Already have an account? Sign in

Assili Salim

Aug 11

Article 50 Is Live. Your Compliance Logs and Your Debug Logs Should Be the Same Thing.

#ai #api #typescript #node

2 min read

Assili Salim

Aug 9

Alert-Based vs. Threshold-Based: Why Monitoring Isn't the Same as Cost Control

#ai #api #llm #typescript

2 min read

Assili Salim

Aug 6

Spot-priced LLM inference changes how runtime budget guards should work

#ai #api #llm #typescript

2 min read

Assili Salim

Aug 5

Model selection isn't just an AI decision. It's a cost decision.

#ai #typescript #agents #node

2 min read

Assili Salim

Aug 4

Migrating from Bedrock Agents Classic to AgentCore? Audit your cost model first.

#ai #typescript #agents #buildinpublic

2 min read

Assili Salim

Aug 3

GPT-5.6 Luna dropped 80% in three weeks. Here's the pricing bug it can introduce into your agent stack.

#ai #api #llm #typescript

2 min read

Assili Salim

Aug 2

Supabase's new agent benchmark reveals 3 lessons production AI agents still ignore

#ai #api #llm #typescript

2 min read

Assili Salim

Jul 31

Two Claude network failures this week. Here's the retry-logic bug they exposed.

#ai #api #llm #typescript

3 min read

Assili Salim

Jul 29

MCP 2026-07-28 shipped today. Here's what it means for agent cost tracking.

#ai #javascript #api #llm

4 min read

Assili Salim

Jul 13

GhostApproval: Why Agent Permissions Need Resolved Runtime State

#ai #javascript #typescript #coding

2 min read

Assili Salim

Jul 12

Agent Telemetry Is Not Agent Control

#ai #javascript #api #llm

2 min read

Assili Salim

Jul 11

AI Agents Need Runtime State Checks, Not Just Better Prompts

#ai #api #llm #agents

4 min read

Assili Salim

Jul 10

GPT-5.6 Ultra Mode Means Agent Budgets Need to Handle Subagents

#ai #developer #tooling #chatgpt

4 min read

Assili Salim

Jul 8

Always-Running AI Agents Need Pre-Call Stop Conditions

#ai #javascript #opensource #api

3 min read

Assili Salim

Jul 4

Tokenizer Changes Can Break AI-Agent Budget Assumptions

#ai #agents #typescript #javascript

3 min read

Assili Salim

Jul 3

Gateway Routing Helps AI Apps. Agent Runtimes Still Need Pre-Call Guards.

#ai #api #llm #agents

4 min read

Assili Salim

Jun 28

Reusable Agent Skills Need Pre-Call Runtime Checks

#ai #javascript #api #llm

4 min read

Assili Salim

Jun 27

If AI Agents Run in Parallel, Budget Checks Need to Happen Before Every Provider Call

#ai #javascript #api #agents

3 min read

Assili Salim

Jun 26

AI Coding Agents Need Runtime Telemetry Before Commit Telemetry

#ai #javascript #api #llm

4 min read

Assili Salim

Jun 25

Usage-Based AI Coding Needs Runtime Budgets, Not Just Billing Dashboards

#ai #api #llm #agents

4 min read

Assili Salim

Jun 24

Claude Had Elevated Errors. Your AI Agent Should Not Turn That Into a Retry Storm.

#ai #api #llm #typescript

3 min read

Assili Salim

Jun 23

AI Coding Agents Need Runtime Guardrails, Not Just Human Review

#ai #api #llm #agents

3 min read

Assili Salim

Jun 22

Why Unknown Model Pricing Should Fail Closed in AI-Agent Runtimes

#ai #api #llm #agents

3 min read

Assili Salim

Jun 21

The AI-Agent Call You Should Block Before It Happens

#ai #api #typescript #agents

3 min read

Assili Salim

Jun 19

Why AI Agents Need Runtime Budgets Before Provider Calls

#ai #api #claude #chatgpt

3 min read

Assili Salim

Jun 18

What is your cutoff for killing a bad Codex run?

#ai #agents #typescript #llm

1 min read

Assili Salim

Jun 17

How to reduce you LLM bill

#ai #api #agents #typescript

1 min read

Assili Salim

Jun 16

I’m building a local-first TypeScript guard for runaway AI-agent costs

#ai #opensource #api #agents

2 min read

DEV Community

Assili Salim

Badges

Writing Debut

Skills/Languages

Currently learning

Currently hacking on

Available for

Multi-Agent Systems Share Context. They Should Share Budgets Too.

Want to connect with Assili Salim?

Article 50 Is Live. Your Compliance Logs and Your Debug Logs Should Be the Same Thing.

Alert-Based vs. Threshold-Based: Why Monitoring Isn't the Same as Cost Control

Spot-priced LLM inference changes how runtime budget guards should work

Model selection isn't just an AI decision. It's a cost decision.

Migrating from Bedrock Agents Classic to AgentCore? Audit your cost model first.

GPT-5.6 Luna dropped 80% in three weeks. Here's the pricing bug it can introduce into your agent stack.

Supabase's new agent benchmark reveals 3 lessons production AI agents still ignore

Two Claude network failures this week. Here's the retry-logic bug they exposed.

MCP 2026-07-28 shipped today. Here's what it means for agent cost tracking.

GhostApproval: Why Agent Permissions Need Resolved Runtime State

Agent Telemetry Is Not Agent Control

AI Agents Need Runtime State Checks, Not Just Better Prompts

GPT-5.6 Ultra Mode Means Agent Budgets Need to Handle Subagents

Always-Running AI Agents Need Pre-Call Stop Conditions

Tokenizer Changes Can Break AI-Agent Budget Assumptions

Gateway Routing Helps AI Apps. Agent Runtimes Still Need Pre-Call Guards.

Reusable Agent Skills Need Pre-Call Runtime Checks

If AI Agents Run in Parallel, Budget Checks Need to Happen Before Every Provider Call

AI Coding Agents Need Runtime Telemetry Before Commit Telemetry

Usage-Based AI Coding Needs Runtime Budgets, Not Just Billing Dashboards

Claude Had Elevated Errors. Your AI Agent Should Not Turn That Into a Retry Storm.

AI Coding Agents Need Runtime Guardrails, Not Just Human Review

Why Unknown Model Pricing Should Fail Closed in AI-Agent Runtimes

The AI-Agent Call You Should Block Before It Happens

Why AI Agents Need Runtime Budgets Before Provider Calls

What is your cutoff for killing a bad Codex run?

How to reduce you LLM bill

I’m building a local-first TypeScript guard for runaway AI-agent costs