skill · Global

ai-science-toolkit

Skills & reviewer agents for AI-first climate science — built and used by a PhD atmospheric scientist

Feature or claim this listing

Represent ai-science-toolkit? Send a claim for review so we can discuss listing updates. Claims do not automatically grant ownership or verification.

Data offered docs-verified: Literature search and synthesis from Zotero, arxiv, bioRxiv, Google Scholar and Consensus (lit-review) · Publication-readiness audit of scientific figures (figure-review) · Senior-engineer repo review with parallel sub-reviews (ai-review) · Stress-testing of claims, results and manuscript sections (reviewer-2) · Listing of running Claude/Codex CLI sessions with resume commands (ai-sessions) · Scaffolding of new repos with a standard structure (repo-init)
Formats docs-verified: Claude Code skills (slash commands, SKILL.md) · Shell scripts (.sh) · Markdown reviewer agents
Access docs-verified: MIT licence. No credentials documented for most of the toolkit; Zotero write access, the Notion work journal and session-name generation need configuration.

Use cases llm-inferred: Literature review to support a flood-risk or drainage design assumption using /lit-review · Publication-style audit of model or monitoring charts before they go into a client report using /figure-review · Stress-testing a flood or water-quality modelling claim for baseline and uncertainty using /reviewer-2 · Senior-engineer review of an in-house analysis repo before handover using /ai-review
Audience llm-inferred: Researchers using Claude Code for scientific computing · Climate and atmospheric scientists · Scientific software developers · Analysts building AI-assisted analysis workflows

Constraints docs-verified: Developed on macOS and Linux; on Windows the shell layer does not run, so skills with shell preambles (handoff, pickup, figure-review, repo-init) fail · Session auto-naming hook works only with iTerm2 or VS Code · Zotero write, Notion work journal and session-name generation need extra configuration · No datasets, gauge feeds or data APIs are described; the toolkit offers skills, reviewer agents and shell utilities only · No UK water, civils or environmental content is described

Tools/install/auth auto-extracted from README; quality score heuristic (stars, recency, tool count, licence).

Metadata observed 16/10/2026 · research sweep: github-search