Convert MLX Model to GGUF (KL Divergence Proof)
mihai
Command Palette
Search for a command to run...
Archive
60 posts published so far
Introducing the MCP Document Processor: A Practical Tool for Document Processing
mihai
The Obedience Trade-off: Why I Swapped Claude Code for Zed + GLM-4.7
mihai
AI Agent Control Protocols: The Missing Link in LLM Engineering
mihai
Qwen3-Coder-Next + MXFP8: The 128GB Local LLM That Runs Predictably
mihai
The Gap Nobody Filled: AI-Powered Workflow Validation for Jira Cloud
mihai
Agent Skills, working: a guide with 6 real Atlassian examples
mihai
Build an LLM-Powered Atlassian Forge App for Jira
mihai
What's New in CogniRunner: Local AI, Zero-Key Models & Actions That Write Back
mihai
The goose swarm, self-verifying: small local models shipping real software
mihai
Inside goose-swarm: How We Turned One Local Model Into a Self-Verifying Fleet
mihai
MLX vs GGUF on Apple Silicon: Benchmarking the Same Local Model Two Ways
mihai
swarm-gym for goose: grading local models by running the code they write
mihai
How to run real end-to-end UI tests against a deployed Forge app (2026)
mihai
Security reports for Forge apps: the clean scan that isn't (2026)
mihai
Making the goose swarm predictable: 602 commits, 100 levers, and three bugs I found writing this
mihai
Your Unattended AI Agent Looks Healthy and Does Nothing: 7 Silent Failures on macOS launchd (2026)
mihai
MXFP8 vs Q8: 10x the weight error, 1% the perplexity
mihai
Hide a Forge module per project: app properties in a display condition
mihai
Claude Code in Zed vs the terminal: what the ACP wire actually carries
mihai
Ling 3.0 Flash won't load on a Mac Studio. Qwen3-Coder-Next does, at 73 tok/s.
mihai
Forge egress: backend ignores your URL path, client enforces it
mihai
jira:fullPage is deprecated: what global:fullPage actually changes for cross-product Forge apps
mihai
goose compaction: my 27B re-read the file 95% of the time — quoting it in the summary didn't help
mihai
MLX memory ceiling on a 96 GB M3 Ultra
mihai