Terminal Coding Agent Optimization
Budget: $25 – $50 USD
We already built a working Terminal Coding Agent using Claude, and now we’re looking for someone who really knows their stuff to help push it to the next level.
This isn’t a beginner gig. We want somebody who’s deep in AI agents, autonomous workflows, terminal automation, and LLM orchestration. Someone who can look at an agent system and immediately spot what’s bottlenecking it.
The agent already works. Now we want to make it smarter, faster, more reliable, and way more capable in real-world coding environments.
What You’ll Be Doing
• Improving autonomous coding + debugging flows
• Making terminal execution more reliable
• Tightening up reasoning and multi step planning
• Improving repo understanding and file editing
• Reducing hallucinations and dumb agent loops
• Building stronger memory + context handling
• Optimizing performance on large codebases
• Helping architect cleaner long term systems
Tech Stack
• Rust
• Claude / OpenAI APIs
• LangGraph / LangChain
• Windows + terminal environments
• Git workflows
• Docker
• RAG / vector DB experience is a big plus
You’re Probably a Good Fit If
• You’ve built AI agents before
• You’ve worked on coding agents or autonomous systems
• You know how to wrangle LLMs without babysitting them every 5 minutes
• You understand real engineering, not just prompt tweaking
• You can actually ship clean systems
Bonus Points If You’ve Touched Stuff Like
• Open Interpreter
• SWE-agent
• Devin style systems
• Cursor workflows
• Aider
• Multi-agent architectures
When Applying
Send over:
• GitHub or portfolio
• What you’d improve first in a coding agent
• Your rate
• Anything cool you’ve built that made you go “yeah this actually cooks”
Looking for somebody sharp. If you know how to build agents that don’t fall apart under pressure, let’s talk.
This isn’t a beginner gig. We want somebody who’s deep in AI agents, autonomous workflows, terminal automation, and LLM orchestration. Someone who can look at an agent system and immediately spot what’s bottlenecking it.
The agent already works. Now we want to make it smarter, faster, more reliable, and way more capable in real-world coding environments.
What You’ll Be Doing
• Improving autonomous coding + debugging flows
• Making terminal execution more reliable
• Tightening up reasoning and multi step planning
• Improving repo understanding and file editing
• Reducing hallucinations and dumb agent loops
• Building stronger memory + context handling
• Optimizing performance on large codebases
• Helping architect cleaner long term systems
Tech Stack
• Rust
• Claude / OpenAI APIs
• LangGraph / LangChain
• Windows + terminal environments
• Git workflows
• Docker
• RAG / vector DB experience is a big plus
You’re Probably a Good Fit If
• You’ve built AI agents before
• You’ve worked on coding agents or autonomous systems
• You know how to wrangle LLMs without babysitting them every 5 minutes
• You understand real engineering, not just prompt tweaking
• You can actually ship clean systems
Bonus Points If You’ve Touched Stuff Like
• Open Interpreter
• SWE-agent
• Devin style systems
• Cursor workflows
• Aider
• Multi-agent architectures
When Applying
Send over:
• GitHub or portfolio
• What you’d improve first in a coding agent
• Your rate
• Anything cool you’ve built that made you go “yeah this actually cooks”
Looking for somebody sharp. If you know how to build agents that don’t fall apart under pressure, let’s talk.