corv89/shannot
View on GitHub ↗ ★ 27 Python archived
Human-in-the-loop execution for LLM agents
7 mentions on Hacker News by 1 person: 0 stories, 7 comments · first 2025-10-24 · latest 2026-01-20
20082026
-
…execution via SSH, checkpoint/rollback for undoing mistakes. Feedback greatly appreciated! [0] https://github.com/corv89/shannot
comment by corv · 2026-01-20 · in Running Claude Code dangerously (safely) -
…working on example scripts for common sysadmin tasks (nginx config, log cleanup, cert audits, etc.) https://github.com/corv89/shannot
comment by corv · 2026-01-11 · in Ask HN: What are you working on? (January 2026) -
…wants to do before approving. WIP, currently porting to PyPy 3.8 to unlock MacOS arm64 support: https://github.com/corv89/shannot
comment by corv · 2026-01-06 · in Sandboxing Untrusted Python -
…in the sandbox, you just see exactly what it wanted to mutate before approving. WIP but core works: https://github.com/corv89/shannot
comment by corv · 2026-01-01 · in Observed Agent Sandbox Bypasses -
…review before executing. It’s still WIP but the core sandbox works. Feedback greatly appreciated: https://github.com/corv89/shannot
comment by corv · 2025-12-30 · in Show HN: A Claude Code plugin that catch destructive Git and filesystem commands -
…on a lightweight approach that augments user namespaces with libseccomp to filter syscalls via BPF. https://github.com/corv89/shannot
comment by corv · 2025-11-11 · in The Paranoid Guide to Running Copilot CLI in a Secure Docker Sandbox -
…go beyond read-only mounts to syscall restrictions. Open to collaboration on pushing this further: https://github.com/corv89/shannot
comment by corv · 2025-10-24 · in Living Dangerously with Claude