Tag AI
- Replaying old code reviews to compare Claude models and effort levels
- PostgreSQL 9.6 to 17 in one go, driven from my phone
- Can a 4B model decide which pull requests need the expensive reviewer?
- Personal Agentic Infrastructure
- How AI agents rekindled passion for my open source projects
- Onboarding packit upstream projects to fullsend (part 1)
- Running opencode inside openshell on Fedora 44 with llama-cpp
- Deploying agents to OpenShift, again (part 2)
- AI tools feel like heroin for productivity
- Deploying agents to OpenShift, again (part 1)
- Backporting code with Claude Code Opus 4.6
- Trying Claude Code prompt injections (Feb 2026)
- Ambient Code followup: PR merged
- Ambient Code: first impressions
- First test of Claude Sonnet 4.5 for an agent that backports a patch for an RPM
- More reliable agents
- Backporting upstream patches with Code Assistants
- Trying Claude Code in July 2025
- Lessons learned from running the Log Detective service
- Comparing llama-cpp and vllm in model serving
- Generating first set of data for LogDetective using InstructLab
- Using InstructLab in Log Detective
- Running logdetective service in containers with CUDA on EC2
- Running logdetective on an EC2 VM with CUDA
- Running logdetective on Red Hat OpenShift AI with CUDA
- Announcing Log Detective