research-paper-writing

Automating the research paper writing workflow involves coordinating experiment design, drafting, revision, and submission. This pipeline uses a SKILL.md file to drive deterministic tasks, manage references, and store artifacts for ML/AI venues like NeurIPS, ICML, and ICLR.

Yes, citation verification and result logging are integrated into the workflow. Automated processes check references while result logging tracks experiment execution, ensuring reproducibility and traceability for your final ML/AI paper submission.

The research paper writing pipeline supports submissions to major ML/AI venues including AAAI, ACL, and COLM. It coordinates the full lifecycle from planning and experiments to drafting, revision, and final submission for these specific conferences.

Structuring directories for a research paper writing pipeline requires optional folders like scripts/, references/, and assets/. Loaded by SKILL.md, these directories perform deterministic tasks, access in-depth references, and store artifacts for the end-to-end workflow.

Start by creating a SKILL.md file with YAML frontmatter containing a name and description. Follow the Phase 0 to Phase 4 workflows to plan, execute, analyze, and draft a submission-ready ML/AI research paper.

Limitations of a modular skill pipeline for ML paper drafting include relying strictly on YAML frontmatter setup and directory structures like scripts/ and references/. Proper artifact storage and deterministic task execution depend entirely on correct SKILL.md configuration.