{"slug": "aftervibe-what-remains-when-the-conversation-ends", "title": "AfterVibe: What Remains When the Conversation Ends", "summary": "Researchers submitted AfterVibe, a framework that recovers natural-language specifications from vibe coding sessions, to arXiv on 10 July 2026. AfterVibe uses an LLM to extract an abstract spec from a code artifact and its conversation trajectory, then validates it by having a second blind AI agent re-implement the artifact from the spec alone and grading the result through a multi-tier pipeline. Across 72 real-world vibe-coded projects from a company's internal coding sessions, multiple independent regenerations scored a mean of 5.06 out of 6.0, and iterative refinement raised the score to 5.74, outperforming existing human-authored descriptions.", "body_md": "# Computer Science > Software Engineering\n\n  [Submitted on 10 Jul 2026]\n\n# Title:AfterVibe: What Remains When the Conversation Ends\n\n[View PDF](https://arxiv.org/pdf/2607.09900)\n\n[HTML (experimental)](https://arxiv.org/html/2607.09900v1)\n\nAbstract:We present AfterVibe, a framework that recovers natural-language specifications from a vibe coding session. Given a code artifact and the conversation trajectory that produced it, AfterVibe uses an LLM to extract an abstract natural-language specification capturing the developer's intent, and validates it through a regeneration test: a second, blind AI agent re-implements the artifact from the spec alone, and the resulting code is graded against the original through a multi-tier validation pipeline. Spec quality is thus measured by whether an agent can regenerate passing code; if the verifiers deem the implementations equivalent the spec is considered strong, otherwise it is iteratively refined. Evaluating AfterVibe on 72 real-world vibe-coded projects from a company's internal coding sessions, we find that its recovered specs are abstract by design-capturing behavioral intent without dictating implementation-yet strong. Multiple independent regenerations achieve a high mean regeneration score of 5.06 out of 6.0 while remaining diverse in their details, confirming that the spec constrains what without over-prescribing how. Besides outperforming existing human-authored descriptions, the specs can be further strengthened iteratively to a score of 5.74. A practical implication is that specifications-not code-could become the primary artifact for human review and the source of record at a time when AI-generated code is outpacing customary code review.\n    \n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer \n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers \n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps \n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations \n\n*(*[What are Smart Citations?](https://www.scite.ai/))\n# Code, Data and Media Associated with this Article\n\nalphaXiv \n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers \n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub \n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub \n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face \n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast \n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))\n# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower \n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender \n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))\n# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.\n\nHave an idea for a project that will add value for arXiv's community? [**Learn more about arXivLabs**](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/aftervibe-what-remains-when-the-conversation-ends", "canonical_source": "https://arxiv.org/abs/2607.09900", "published_at": "2026-09-29 02:39:32+00:00", "updated_at": "2026-09-29 03:17:52.621176+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-agents", "ai-research", "developer-tools"], "entities": ["AfterVibe", "arXiv"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/aftervibe-what-remains-when-the-conversation-ends", "markdown": "https://wpnews.pro/news/aftervibe-what-remains-when-the-conversation-ends.md", "text": "https://wpnews.pro/news/aftervibe-what-remains-when-the-conversation-ends.txt", "jsonld": "https://wpnews.pro/news/aftervibe-what-remains-when-the-conversation-ends.jsonld"}}