An agent skill that strips AI tone from Chinese text, built on a corpus study
Emerged finds new products worth a look, daily — what’s new today
It fixes the causality, emotion and venue mismatches that make AI prose read as AI, so teams shipping agent-written docs and fiction get output people will actually read.
Sepia is an open-source writing skill for AI agents that treats de-AI-ing as a structural problem rather than a word swap. For fiction it repairs theme, causality, how emotion is rendered, and character networks. For professional prose it applies venue-specific rules to release notes, pull requests, postmortems and more. It offers four operations: write, review (diagnose only), refactor (minimal edits) and recreate (full rewrite). Voice profiles such as a built-in Hemingway stack on top. It installs via the Skills CLI for 77+ agents, with native plugins for Claude Code and Codex.
An agent skill that strips AI tone from Chinese text, built on a corpus study
An agent skill that edits LLM prose toward Simplified Technical English
Your coding agent hands back a built page instead of a wall of prose
A writing page that keeps a tamper-evident log of how each piece was typed
A tiny native terminal agent where you co-write the spec before any code
Zhipu's open-source workbench for running AI agents on a codebase