Template repo with agent's configuration for working in scientific papers
Rezenders/scientific_writing_agent
Skill Claude CodeCodex
Use when a reviewer comment (human or agent) targets a specific paragraph or section and requires a written response that modifies manuscript text. Never modifies the file without explicit user approval.
Rezenders/scientific_writing_agent
Skill Claude CodeCodex
Use at the start of a new paper project, or after a long break, to pre-populate all agent memory files with manuscript-specific knowledge extracted from CLAUDE.md and the manuscript itself. Run this once after setup to give agents immediate context rather than building it up slowly over sessions.
Rezenders/scientific_writing_agent
Skill Claude CodeCodex
Use after making targeted edits to one or more manuscript files to review only what changed — not the whole section. Runs consistency and scientific review scoped to the diff, checking for new inconsistencies introduced, claims that now need evidence, and prose that broke from surrounding context. Much faster than…
Rezenders/scientific_writing_agent
Skill Claude CodeCodex
Use when you need an inventory of all open inline annotations in the manuscript — todos, revision markers, and author comments — before a writing session or submission.
Rezenders/scientific_writing_agent
Skill Claude CodeCodex
Use when preparing for submission or after a major revision cycle to run a complete audit of the manuscript across all sections. Orchestrates consistency, scientific review, and (optionally) implementation checking in a structured multi-phase workflow. Use before /pre-submission to catch issues early.
Rezenders/scientific_writing_agent
Skill Claude CodeCodex
Build the manuscript PDF and run chktex lint. Reports LaTeX errors and warnings. Run after any substantive manuscript edit.
Rezenders/scientific_writing_agent
Skill Claude CodeCodex
Use when preparing the manuscript for submission. Runs a structured go/no-go checklist covering build, lint, open annotations, consistency, and (optionally) implementation alignment.
Rezenders/scientific_writing_agent
Skill Claude CodeCodex
Use when preparing a formal response letter for a journal or conference resubmission. Takes the full review (all reviewers, all comments) and produces a structured response letter with per-comment responses and manuscript edit proposals. Nothing is written to file without explicit user approval. Differs from…
Rezenders/scientific_writing_agent
Skill Claude CodeCodex
Use when a full section has been written or substantially rewritten and needs a combined scientific review and consistency check against the rest of the manuscript.
Rezenders/scientific_writing_agent
Skill Claude CodeCodex
Use when pulling from or pushing to a remote git repository (e.g. Overleaf, GitHub, or any other remote). Enforces safe pull-before-push, stages only manuscript source files (never build artifacts), and uses a consistent commit message format.
Rezenders/scientific_writing_agent
Skill Claude CodeCodex
Use when a manuscript paragraph describes implementation behavior (functions, interfaces, data structures, system methods) and needs accuracy verification plus rewrite suggestions. Never modifies text without explicit user approval. OPTIONAL: only useful if your paper has an associated software implementation.
Rezenders/scientific_writing_agent
Skill Claude CodeCodex
Use when you want to write or substantially rewrite a manuscript paragraph or short passage toward a stated goal — a framing, claim, or rhetorical function the passage should accomplish. Not for minor edits; not for responding to a specific reviewer comment.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: