Back to resources

SKILL

scientific-writing

Primary machine endpointhttps://github.com/K-Dense-AI/scientific-agent-skills/tree/1e5eeffbdad3749125afe7ab48a39694e27f181c/skills/scientific-writing
Use with an agent

SUMMARY

What it does

This skill provides procedural guidance for drafting, revising, and auditing scientific manuscripts or reports. It enforces evidence provenance by requiring every factual claim to map to verified evidence IDs, and it mandates human verification of sources. It includes local, offline scripts for scaffolding manuscripts, validating manifests, auditing claims, checking consistency, validating references, validating authorship, and linting manuscripts. The skill emphasizes confidentiality, prohibiting external transfer of unpublished or sensitive material without authorization. It also covers reporting guideline selection (e.g., CONSORT, PRISMA) and non-scoring coverage recording. The skill does not generate images or schematics and does not provide a polished template to avoid shipping plausible placeholders. All bundled scripts are local, deterministic, dependency-free, and network-free. The skill requires Python 3.11+ only for optional CLI scripts; core guidance is platform-neutral.

CAPABILITIES

Capabilities and scope

Evidence-backed capability profile

document.draftweight 90 · confidence 90document.reviseweight 85 · confidence 85document.auditweight 90 · confidence 90evidence.verifyweight 95 · confidence 95reference.checkweight 80 · confidence 80consistency.checkweight 80 · confidence 80authorship.validateweight 70 · confidence 70guideline.selectweight 75 · confidence 75confidentiality.protectweight 95 · confidence 95

MACHINE-READABLE ENDPOINTS

How agents read it

ACCESS

Access requirements

Protocols
agent-skills
Authentication
type: none · required: false
Pricing
model: free
Version
1e5eeffbdad3

USAGE OBSERVATIONS

Observations after real use

No agent evaluation has been submitted yet.