
Herdr plugin: stash prompts for an agent ahead of time and insert them when it's ready
Pouch is a Herdr plugin for stashing prompts before an agent needs them and inserting them when it is ready. It lets users queue instructions while an agent is still working and hand them over one keypress at a time. Text is typed into the input, never submitted, so the user always presses Enter. The plugin requires Herdr 0.7+ and Node 22.6+ or Bun, with no build step or clone. Setup writes one marked block to ~/.config/herdr/config.toml, and uninstall keeps stashed messages because they are the user's writing, not a cache. Users can stash from a shell, a script, or one agent preparing work for another. The CLI includes add, list, show, insert, rm, clear, strip, unstrip, open, browse, sync, key, setup, update, and uninstall. A popup supports moving, filtering, copying, editing, deleting, reordering, switching pouches, and choosing which pane inserts go to. Storage uses one JSON file per pouch under pouches/, keyed by workspace label, cwd, and agent, never by pane id. Every write takes a lock and re-reads inside it.
Teams use Pouch to queue the next instructions while an agent is still working.
Teams use Pouch to stash prompts from a shell, a script, or another agent.
Teams use Pouch to insert a stashed message into an agent's input without submitting it.
Teams use Pouch to filter messages as they type in the popup.
Teams use Pouch to copy a message to the clipboard.
Teams use Pouch to switch between pouches on disk, including orphaned ones.
Teams use Pouch to pin a three-row strip under a pane.
Teams use Pouch to remove a message or empty the pouch.
Lynote combines AI-draft rewriting and AI-text detection with note-taking, transcription, video summaries, and study tools.
LMSYS Chatbot Arena is a crowdsourced open platform for LLM evals. Collected over 1,000,000 human pairwise comparisons to rank LLMs with the Bradley-Terry model and display the model ratings in Elo-scale.
Zhipu Qingyan is a GLM-based AI assistant whose official description emphasizes understanding goals, breaking down tasks, and using tools.
字节跳动AI助手