A developer built board.md, a local kanban board that reads and edits markdown task files already stored in a repository, after finding that coding agents only needed readable text files rather than a…
A developer built a PreToolUse hook to block destructive Bash commands from an AI coding agent, but the initial blocklist failed four self-authored bypasses — base64 pipes, python3 -c, find -delete, a…
A developer built Ordewell, an Apache 2.0 tool that gives each parallel coding-agent task its own git worktree so concurrent sessions cannot silently overwrite each other's files. The system creates w…
A developer documented how to fork Claude Code and Codex agent sessions so an alternative approach can be tried without polluting or losing the original conversation context. Claude Code offers /branc…
Engineer Derek Wang argues that AI coding harnesses should be built as Popperian refutation machines rather than pass/fail gates, citing a full regression suite that runs in ten seconds. His `fulltest…
Meta's CLI coding agent Muse Code, launched in August 2026 on Muse Spark 1.2, uses a parent/child architecture in which each write-capable child agent gets its own git worktree rather than sharing a s…
A developer tested Claude Code 2.1.278's Bash(git push:*) deny rule against 14 different spellings of a git push command and found the rule blocked only 8 of them, with 5 variants successfully reachin…
A developer found that Claude Code 2.1.263 silently deletes a clean, unnamed worktree and its branch on /exit without prompting, running git worktree unlock, git worktree remove --force, and git branc…
A developer has described a workflow for running up to four Claude Code agents in parallel on a single repository by giving each agent its own git worktree, a scoped task contract, and a serial merge …
Developer Vishal Narkhede open-sourced Agentdock, a web dashboard for managing parallel AI coding agents like Claude Code and Cursor Agent across multiple repositories via tmux sessions with git workt…
A developer's comparison of Notion versus a markdown-based stack finds that markdown setups win on wikis, offline use, history, and AI integration, while Notion wins on real-time co-editing, permissio…
Alex Shev's comment inspired a NestMux developer to detail the practical challenges of running coding agents in parallel on Windows, focusing on isolation, file ownership, and cleanup. The post explai…
A developer's tests show that Claude Code's internal configuration layers fail to prevent the coding agent from writing files, while a macOS sandbox profile consistently blocks it. Across measurements…
Anthropic fixed a Claude Code sandbox escape in version 2.1.247 after researchers at Accomplish showed that an untrusted repository could run commands on a Mac outside the macOS Seatbelt sandbox with …
A solo developer who runs an AI-powered company using Claude Code to write and maintain code has documented two incidents where an AI agent caused problems when left unattended, leading to the creatio…
A developer discovered that their macOS backup job had silently failed for three weeks because launchd couldn't write to ~/Documents due to macOS TCC permissions. The developer solved the issue by mov…
OpenCode, a terminal AI agent running the 200k-context open-weight LLM big-pickle, enabled a user with 15 years of git experience to build more real automation in two days than in most of that time, b…
A developer warns against trusting AI-generated code refactors until the observable output matches the original contract, highlighting a case where a patch dropped a required 'apiVersion' field from a…
Syntaf released hop v0.1.0, a CLI tool that searches all git repositories under a user's code directory and jumps to the worktree holding a given branch, with tiered matching, tab completion, and an i…
A developer discovered that the commit hook in their repository, which uses Claude CLI to generate commit messages, has a guard clause that prevents it from running on commits made with `git commit -m…