Show HN: Stash, turn your Instagram saves into notes Claude finds on its own
Stash is an open-source tool that converts Instagram saved posts into searchable notes, with Claude surfacing relevant ones automatically.
Stash, an open-source project by developer ParthAkholkar, launched on August 22, 2026, via a Show HN post. The tool takes a user's Instagram saved posts and converts them into notes that Anthropic's Claude AI model can retrieve on its own, without manual search.
Per the GitHub repository, Stash pulls from Instagram's saved content and structures it into a note format that Claude can index and surface contextually — the pitch being that saves users already make on Instagram stop becoming a graveyard and start functioning as a retrievable knowledge base.
The project is the work of a single developer, ParthAkholkar, and was posted under the "Show HN" tag on Hacker News, a typical venue for early-stage, indie-built tools seeking community feedback. No funding, team size, or commercial entity is mentioned in the available source material. The repository is publicly available on GitHub, suggesting an open-source, self-hostable approach rather than a hosted SaaS product.
Stash sits at the intersection of personal knowledge management (PKM) tooling and AI-assisted retrieval — a space that has drawn a number of tools in recent years, including Readwise, Notion AI, and various "second brain" apps. The specific angle here — Instagram saves as input, rather than articles, PDFs, or web clips — is less common, targeting content that users bookmark on a social platform but rarely return to. Claude's role appears to be agentic retrieval: finding notes without the user needing to query explicitly, per the project's own framing.
At the time of publication, the Show HN post had 1 point and no comments, per Hacker News, indicating it is very early in community exposure. No roadmap, expansion plans, or next steps are stated in the available source material beyond what the repository itself documents.
No comments yet — start the thread.