Launch HN: Vespper (YC F24) ships SOTA Docx MCP, claims 3x faster editing
topaztee · hn · 2026-09-29
Vespper (YC F24) launched a SOTA Docx MCP letting AI agents efficiently edit Word documents, powered by a fine-tuned model — claiming 3x faster, 2x cheaper, and more accurate than the closest alternative.
Key points:
- The team built an AI regulatory document editor for pharma; users wanted their own Word templates, exposing how bad agents are at OOXML.
- Word files are zips of verbose XML: adding a numbered list requires a new numbering.xml entry linked back into document.xml; bolding splits text into 3+ run elements — burning agent time and tokens.
- Existing approaches all fall short: raw zip editing, low-level libraries or opinionated MCPs (SuperDoc, Office CLI, Adeu), or Markdown/HTML round-trips via pandoc/mammoth.js.
Team: Topaz was a senior SWE at Snyk; Dudu a DL engineer at Viz.ai.
More from coding & agent
- Google Research open-sources PaperOrchestra, a multi-agent framework for writing AI papers — google-research · 2026-11-16
- ffmpeg-skill gives coding agents local video editing via 42 FFmpeg tools and MCP — kajisho5 · 2026-11-16
- Agents Now Reply to Slack Messages on Their Own, Resolving a Prod Incident Ghost-to-Ghost — iamrobotbear · 2026-11-16
- Engineer's Interview Prep Method: Use Claude to Find Which Answers Undersell Your AI Work — brandon_galang · 2026-11-16
- DeepSeek open-sources its app frontend "chat-demo", already at 3.9k stars — NielsRogge · 2026-11-16
- lemire: AI's real impact on coding is expanding the engineer's 'cognitive radius', not speed — lemire · 2026-11-16