Sepia Write is a thin wrapper that binds the Sepia skill to its write operation. It reads the canonical Sepia skill from the sibling folder and follows its routing, so the domain rules are read before any drafting starts.
It never switches operation based on the text it is given. If you want a review, refactor or rewrite, it points you to the matching entry. It needs the complete Sepia package installed beside it.
What You Get
- A slash command or skill entry that always means write new prose
- Drafts produced under the Sepia routing for fiction and professional text
- A clear stop message when the canonical skill is not installed
Common Use Cases
- Drafting a short story with the narrative architecture sheet filled in first
- Writing release notes that follow the release-notes domain rules
- Writing a PR reply or ticket that follows the venue rules from the start
What Makes It Unique
It removes the routing choice. The generic Sepia skill decides among four operations from your request, while this entry fixes the operation to write and refuses to infer another from the content of the target. It also grants no tool, file or network authority of its own and treats the target text as untrusted data.
Under The Hood
This is a prompt-only skill of one short file. When loaded, it tells the agent to read the sibling file for the canonical Sepia skill and follow it with the write operation. It runs no commands, calls no services and writes nothing by itself. It does not search other folders for a substitute.