HN Explorer

CuriouslyC

↗ HN profile [ 60.0 ms ] full profile
Karma
0
Created
()
Submissions
0
  1. I Changed My Mind About MCP (sibylline.dev)
    3 points 0
  2. You Don't Need to Detect Prompt Injection to Stop It (sibylline.dev)
    2 points 0
  3. Smith: The Secure Open Source Multi-User AI Assistant Framework (github.com)
    4 points 2
  4. Defeating Prompt Injection with Protocol Firewalls (sibylline.dev)
    2 points 0
  5. The Pillars of Agent Security (sibylline.dev)
    2 points 0
  6. The Pillars of Agentic Security (sibylline.dev)
    1 points 0
  7. Clean: High Performance Prompt Injection Detection and Mitigation (github.com)
    1 points 1
  8. GLM-5: Targeting complex systems engineering and long-horizon agentic tasks (z.ai)
    484 points 94
  9. Scurl: Agent First Curl Wrapper with Markdown Extraction and Secret Blocking (github.com)
    1 points 1
  10. Pi Is the Linux of Agent Harnesses (sibylline.dev)
    2 points 0
  11. Averting the Code Quality Apocalypse (sibylline.dev)
    3 points 0
  12. Stop screwing around with agent orchestration, your bottleneck is validation (sibylline.dev)
    2 points 1
  13. The Problems with Spec Driven Development (sibylline.dev)
    1 points 0
  14. Agents with Scribe Solve Hard Problems 57% More Often (sibylline.dev)
    1 points 0
  15. Stop screwing around with agent orchestration, your bottleneck is validation (sibylline.dev)
    2 points 1
  16. Scribe Increases Agent Success Rate on Hard Problems by 57% (sibylline.dev)
    1 points 0
  17. Scribe reduces agent token usage by 30% with no loss of accuracy (sibylline.dev)
    2 points 0
  18. Scribe reduces SWE-bench token usage by 30% with no loss of accuracy (sibylline.dev)
    1 points 0
  19. Is your codebase holding back your AI tools? (codehealth.sibylline.dev)
    1 points 1
  20. Digital Alchemy: Turning Slop into Gold with Ralph and Valknut (sibylline.dev)
    1 points 0
  21. The Code Quality Leaderboard (codehealth.sibylline.dev)
    1 points 0
  22. Model Evals Can Fix Education (sibylline.dev)
    2 points 0
  23. Valknut: Code Intelligence for Agents (sibylline.dev)
    2 points 0
  24. Valknut: Advanced Code Intelligence (sibylline.dev)
    1 points 0
  25. Show HN: Valknut – static analysis to tame agent tech debt (github.com)

    Hi y'all, In my work to reduce the amount of time I spend in the agentic development loop, I observed that code structure was one of the biggest determinants in agent task success. Ironically, agents aren't good at…

    2 points 1