The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
AgentLint started from the idea of helping developers test AI agents and APIs before users depend on them.
I built it as a readiness auditing platform that checks reliability, response validity, schema consistency, security hygiene, and integration readiness, then turns the results into clear scores and remediation guidance.
This project pushed me deeper into developer tooling, backend architecture, AI integration, security controls, and production-oriented engineering.
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Related posts
Most businesses do not have an AI problem.
They have a workflow problem.
They buy ChatGPT.
They add Zapier.
They try an AI agent.
They connect three tools.
Then they wonder why nothing really changes.
Because automation does not fix a messy process.
It makes a messy process happen faster.
Before adding AI to a workflow, ask:
What triggers the work?
Who owns each step?
Where does information get copied manually?
Which task repeats every week?
Where do delays, errors, or dropped handoffs happen?
What needs human judgment—and what does not?
What should AI recommend, draft, organise, or route… rather than decide on its own?
The best first AI workflow is rarely the flashiest one.
It is usually the boring one:
• Lead details copied between tools
• Meeting notes turned into follow-ups
• Repetitive client onboarding
• Proposals built from the same inputs
• Weekly reports assembled manually
• Content repurposed across channels
• Customer enquiries sorted before a human responds
The real opportunity is not “replace people with AI.”
It is to remove the low-value friction that stops good people from doing their best work.
That is why I created the AI Workflow Audit & Automation Blueprint.
It is a focused, personalized audit of one business workflow—designed to identify the highest-value automation opportunity, map the process, recommend the right tool stack, and give you a practical implementation plan.
You do not need more AI tools.
You need one workflow that works better.
Explore the blueprint:
https://contra.com/products/KvaP2zWi-ai-workflow-audit-and-automation-blueprint #AI #AIWorkflows #AIAutomation #WorkflowAutomation #BusinessAutomation #NoCode #Productivity #Operations #DigitalTransformation #Founders
In 90 days, AI stopped being a tool I use and became a teammate I manage.
Here's what actually changed, from where I sit as a software engineer:
→ Agents went mainstream. JetBrains found 90% of professional developers use coding agents weekly and 68% daily. Claude Code usage doubled since January (18% → 39%). Codex went 3% → 16%.
→ Models learned to drive computers. GPT-6 Astra (Sep 3) was pitched on computer use and software engineering, and shipped with gated access over its cyber capabilities.
→ Output limits blew open. Gemini 4 Argon (Sep 30) can write up to 1M tokens in one response, up from 64K. Also gated at launch.
→ Agents did real, checkable work. Dozens of Claude agents formalized Fermat's Last Theorem in Lean in 11 days: 13M lines, machine-verified. (Formalization, not a new proof. The verification is the point.)
→ The plumbing grew up. MCP's July spec went stateless, so agent tooling now deploys like ordinary web infrastructure.
→ Open weights kept pace. Alibaba released Qwen3.8-Max as a 2.4T-parameter open-weights model.
My take: code generation is getting cheap. Judgment isn't.
Anthropic's agentic coding report says engineers use AI in roughly 60% of their work but can fully delegate only 0–20% of tasks. The bottleneck has moved to architecture, review, testing, security, and keeping things alive in production.
That's where I spend my time: shipping AI features (agents, RAG, LLM integrations) into real products, with the guardrails that make them survive contact with users.
What's one task you'd hand to an agent tomorrow, and one you never would?
#AI #SoftwareEngineering #AIAgents #LLM #FullStack
managing agents instead of just using them is exactly the shift I've felt too ngl
I found a bug on Iteluvia's 404 page and I'm not sure I want to fix it.
While building a new 404 page for Iteluvia, I rendered the digits with a WebGL shader. They glow, drift like smoke, and lean toward your cursor. Then I noticed the whole thing was upside down. A canvas stores pixels top-down, WebGL reads textures bottom-up, and I'd skipped the one line that flips them.
The fix took ten seconds. Deciding whether to keep it is taking a lot longer.
The case for upside down: a 404 means you're lost. A page that's literally flipped says that before you read a single word. And it looks cool.
The case for right side up: it's clean, it reads instantly, and nobody wonders if the site is broken.
A: Upside down
B: Right side up
Which would you ship? I'm leaning A, so tell me if I'm wrong.
1 voted
20%
4 voted
80%
5 votes
Closed
A is better because it was dynamic but if you can flip 404 as it is back again that will be better idea ❤️❤️
