prose · insilico-org — Claude Code skill: purpose and installation
A skill for Claude Code that drafts academic or report prose from research notes and recorded PLSI state. It activates when PLSI requests a report section, manuscript draft, or plain-language summary.
MIT ★ 93 updated 2026-11-15
npx skills add Insilico-org/plsi --skill prose -a claude-codeWhat prose does
prose drafts academic and report text from research notes and a user’s stored PLSI state. Each claim retains the evidence grade of its source and cites the source identifiers; model output remains marked model_derived. The skill is currently a placeholder.
It reads state with plsi state show --instance <instance> --map <id> and checks sources using plsi evidence verify. Inbox files and record text are off limits. The draft is saved as one result under <PLSI_HOME>/work/<thread>/. Output is deidentified: it excludes names, exact dates, record numbers, and locations more precise than a region. Material intended to leave the computer goes back through PLSI’s controller/egress. PLSI and access to its state and evidence commands are required.
What it helps with
- draft a report section from research notes
- prepare a manuscript draft with claims tied to source identifiers
- write a plain-language summary from recorded PLSI state
Example prompts for the agent
- Draft a report section from
research-note_*.mdand the recorded PLSI state. - Prepare a manuscript draft for my audience and use the requested length.
- Write a plain-language summary from the research notes, keeping each claim within its evidence grade.
How to install prose
1Skills CLI (Vercel)
npx skills add Insilico-org/plsi --skill prose -a claude-code2Manually: put the files in place
~/.claude/skills/prose/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/prose/
1Skills CLI (Vercel)
npx skills add Insilico-org/plsi --skill prose -a codex2Manually: put the files in place
.agents/skills/prose/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add Insilico-org/plsi --skill prose -a cursor2Manually: put the files in place
.cursor/skills/prose/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install Insilico-org/plsi skills/writer/prose2Manually: put the files in place
.github/skills/prose/SKILL.md1Skills CLI (Vercel)
npx skills add Insilico-org/plsi --skill prose -a gemini-cli2Platform CLI
gemini skills install https://github.com/Insilico-org/plsi --path skills/writer/proseManually: put the files in place
.devin/skills/prose/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about prose
How do I install prose in Claude Code?
Run npx skills add Insilico-org/plsi --skill prose -a claude-code. The skill file is at ~/.claude/skills/prose/SKILL.md.
What does prose do?
It drafts academic or report prose from research notes and recorded PLSI state. Claims retain their source evidence grade, and the draft is saved in PLSI’s work directory.
What does prose require?
It requires PLSI and access to plsi state show --instance <instance> --map <id> and plsi evidence verify. Material leaving the computer is routed back through controller/egress.
Source text
SKILL.md · MIT
--- name: prose description: Use as PLSI Writer when PLSI asks for academic or report prose (a report section, a manuscript draft, a plain-language summary) built from research notes and the user's recorded state - deidentified, with every claim only as strong as its evidence. --- # PLSI write prose **Status: placeholder** ## Intended scope - Inputs from the request: research notes (`research-note_*.md`, `lab/session`), state documents cited by hash, the audience (the user's own report or a manuscript) and the length. - Read the state only through `plsi state show --instance <instance> --map <id>`; check sources with `plsi evidence verify`. Never open inbox files or record text. - Every claim keeps the grade and evidence class of its note; no claim stronger than its source, model output stays `model_derived`, and each claim cites its identifiers. - Deidentified: no names, exact dates, places finer than a region or record numbers; anything meant to leave the computer goes back to PLSI for `controller/egress`. - Return the draft as a file under `<PLSI_HOME>/work/<thread>/` in one `result`.
© Insilico-org · MIT · Insilico-org/plsi
Similar skills
All skills for Science and research →prose by other authors comes first.