ds-lite-analysis-write

Converts reviewed research evidence into bounded claims, analyses, and paper sections.

2|Updated Jun 18, 2026
One-click install
npx skills add https://github.com/AlexenderSokolov/deepscientist-lite-codex-plugin --skill ds-lite-analysis-write-alexendersokolov
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ds-lite-analysis-write
Source: https://github.com/AlexenderSokolov/deepscientist-lite-codex-plugin/tree/main/plugins/deepscientist-lite-core/skills/ds-lite-analysis-write
Command: npx skills add https://github.com/AlexenderSokolov/deepscientist-lite-codex-plugin --skill ds-lite-analysis-write-alexendersokolov

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Research teams often turn experimental results into claims without verifying that the evidence actually supports them, producing overstated conclusions and unrecoverable handoffs. This Skill enforces a review-gated workflow so that only evidence passing a typed review can become claims, limitations, or paper sections. ## Core Features & Use Cases - Evidence-Gated Claim Writing: Builds a claim table linking each claim to its supporting artifact, metric, review lane, limitation, and confidence before any prose is written. - Review Verdict Enforcement: Routes experiments without a valid typed review to the review skill, blocks promotion of failed or needs-human verdicts, and preserves negative or inconclusive results as valid analyses. - Polishing and Writing Routing: Delegates existing manuscript prose to the Nature polishing workflow and new manuscript drafting to the Nature writing workflow when the Academic package is installed, with an evidence-preserving fallback. - Use Case: After an experiment passes review, use this Skill to write an analysis document that states the hypothesis, change, result, tradeoff, and next candidate, then record the analysis node in the research graph and produce a recoverable handoff. ## Quick Start Ask the assistant to summarize only the reviewed evidence, its limitations, and the next defensible step for the current experiment.

Frequently Asked Questions about ds-lite-analysis-write

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I turn experiment results into research claims?▼

Run this Skill after the experiment passes a typed review. It builds a claim table linking each claim to its supporting artifact, metric, limitation, and confidence, then writes an analysis document following the hypothesis-change-result-tradeoff structure.

What happens if my experiment review failed or is missing?▼

A missing review routes the experiment to the review skill first. A fail or needs-human verdict blocks claim promotion entirely; only the limitation and required follow-up are recorded, never polished success prose.

Can I use this to polish an existing manuscript paragraph?▼

Yes, when facts and citation intent are already supplied, it routes to the Nature polishing workflow from the Academic package. If that package is unavailable, it applies an evidence-preserving fallback and reports that the Nature route was not observed.

Does this skill handle negative or inconclusive results?▼

Yes, inconclusive and refuted claim assessments remain valid negative analyses and are preserved rather than rewritten. The workflow prohibits upgrading weak evidence into strong claims or hiding unresolved human checks.

Why does the skill refuse to write my conclusion?▼

It blocks writing when no passing typed review exists, when the claim assessment is not supportable, or when requested wording would conceal missing evidence. Resolve the review gate or adjust the claim to match the evidence first.