tnl - typed natural language plan mode
Persistent plan-mode for coding agents: agents propose a typed English contract (paths, MUST/SHOULD/MAY behaviors, non-goals), you approve, and every future session reads it before coding.
This entry doesn't have a long-form writeup yet. Follow the source link above for the full context.
Featured in
Claude Code tools, plugins, and integrations
The best tools, MCP servers, and harnesses for getting more out of Claude Code - orchestration, observability, telemetry, and remote control.
Tools for OpenAI Codex CLI
The Codex-aware slice of the directory: orchestration, observability, sandboxes, and bridges built specifically for the OpenAI Codex runtime.
AGENTS.md and agent config tooling
Templates, editors, and cross-tool managers for the AGENTS.md ecosystem and adjacent agent config files.
Related entries
agent-style - 21 writing rules for coding agents
Drop-in style guide for Claude Code, Codex, Copilot, Cursor, and Aider that makes their prose and commit messages read like a senior engineer instead of an LLM.
IDEA.md - portable idea-intent file for agents
Standard markdown format for capturing the intent of a feature so any agent (Claude Code, Codex, Gemini CLI, Copilot) can build it. Vendor-neutral counterpart to AGENTS.md.
agents-md - drop-in AGENTS.md template
Curated AGENTS.md preset that kills sycophancy, blocks drive-by refactors, and forces verification loops. Synthesizes Karpathy's principles with Cherny's Claude Code workflow.
anywhere-agents - portable multi-agent config
Single config that follows you across every project and session: curated writing/routing/skills, plus a destructive-command guard for Claude Code, Codex, and friends.