2025-07-18 Fri

• #Wordsmithing

  • #Updated
    • AI Output Post-Formatting

• Py/Servers

  • #Filed
    • Py/Server/ASGI
      • uvicorn
    • Py/Logging/How To/Observe uvicorn and FastAPI app logging to verify JSON logs

• Markdown

  • #Filed
    • Markdown/Idea/25/07/Metalibrary to enable schema to markdown syntax formatting using LLMs
    • Deepdwn - an offline-first markdown application for desktop

• #AI

  • #Docs

    • DeepWiki

  • LLM Concerns

    • AI/LLM/Problem/Context Limit
      • CursorAI/Forum/25/03/Whole 200k Context Window Claude 3.7 Sonnet Max
        • The "Whole 200k Context Window" of Claude 3.7 Sonnet Max - Feedback - Cursor - Community Forum

          • AI/Paper/25/01/LongProc Benchmarking long-context LLMs on procedural generation

            • 70k input 70k output 60k reasoning = 200k, a simple refactor job or documentation request will get to 200k
              

            • 
          • InftyThink: Breaking the Length Limits of Long-Context Reasoning in Large Language Models
            • 

  • AI Coding

    • Person/Asgeir Thor/GitHub/system_prompt_leaks
    • AI/Coding/Problem/Environment Reproducibility
      • Dagger

    • Claude/Code

      • #Filed
        • Person/Michael Yong/Blog/25/07/Claude Code Unleashed
          • Claude/Code/Tool/Claude Code Terragon
            • Terragon - Background Agents for Claude Code
              • Run coding agents in parallel inside remote sandboxes. Automate multiple development tasks concurrently with AI-powered assistance
        • Person/AICodeKing/YouTube/25/06/Claudia - overview of Anthropic/AI/App/Claude Code/Tool/Claudia, a new GUI wrapper
        • Claude/Code/How To/Export Conversation to File
    • #Updated
      • Person/Brian Madison/GitHub/BMAD-METHOD
    • #Filed
      • Person/Sean Grove/YouTube/25/07/The New Code
        • • My Notes
            • In particular, I'm interested in reading the Deliberative Alignment paper he mentioned at 13:48, where an AI/Model/Spec ( #Example: OpenAI/GitHub/model_spec (what I have been calling an AI/Behavior/Catalog)) would be used by a more powerful LLM-as-Judge to score the input output pair, which then was fed back into fine-tuning.

• #Observability

  • #Updated
    • OpenTelemetry/Research/AI/25/07/OTel Logs in FastAPI Context