The latest articles on DEV Community by OxainZ (@oxainz). https://dev.to/oxainz https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3840075%2Fd4a028f9-3278-4c92-8ee0-f0c4d6eac415.png https://dev.to/oxainz en OxainZ Mon, 23 Mar 2026 15:20:42 +0000 https://dev.to/oxainz/i-built-a-vs-code-extension-that-gives-any-ai-your-full-codebase-in-one-click-4fjh https://dev.to/oxainz/i-built-a-vs-code-extension-that-gives-any-ai-your-full-codebase-in-one-click-4fjh <p>Every time I started a new Claude session I was manually re-uploading the same files over and over. Hit the 20-file limit constantly. Lost context mid-build. Got fed up.<br> Built a VS Code extension called SendToAI. One click bundles your whole project into a single clipboard paste — paste it into Claude, ChatGPT, Gemini, whatever and it instantly has your full codebase.<br> The token stuff is actually useful — it shows you exactly how much you’re sending before you send it:<br> ∙ Live token count as you pick files<br> ∙ Cost estimate per model (Haiku vs Sonnet vs Opus) so you’re not burning money blindly<br> ∙ Only 6% of Claude’s 200k context window for a typical project<br> ∙ Compact mode strips comments and saves ~20% tokens automatically<br> Also has project notes that persist between sessions — your stack, architecture decisions, whatever — so Claude always has context without you typing it again.<br> The meta moment: hit my context limit building the extension itself, used it to bundle its own source, pasted into a new chat and kept going. Didn’t lose a single thing.<br> Free to install — search SendToAI in VS Code or here:<br> <a href="https://marketplace.visualstudio.com/items?itemName=OxainZ.sendtoai" rel="noopener noreferrer">https://marketplace.visualstudio.com/items?itemName=OxainZ.sendtoai</a><br> GitHub:<br> <a href="https://github.com/OxainZ/sendtoai-vscode" rel="noopener noreferrer">https://github.com/OxainZ/sendtoai-vscode</a><br> Would love feedback from anyone who tries it.​​​​​​​​​​​​​​​​</p> typescript vscode opensource developers OxainZ Mon, 23 Mar 2026 14:15:05 +0000 https://dev.to/oxainz/i-built-a-vs-code-extension-that-gives-any-ai-your-full-codebase-in-one-click-1kj3 https://dev.to/oxainz/i-built-a-vs-code-extension-that-gives-any-ai-your-full-codebase-in-one-click-1kj3 <p>Every time I started a new Claude session I was manually re-uploading the same files over and over. Hit the 20-file limit constantly. Lost context mid-build. Got fed up.<br> Built a VS Code extension called SendToAI. One click bundles your whole project into a single clipboard paste — paste it into Claude, ChatGPT, Gemini, whatever and it instantly has your full codebase.<br> The token stuff is actually useful — it shows you exactly how much you’re sending before you send it:<br> ∙ Live token count as you pick files<br> ∙ Cost estimate per model (Haiku vs Sonnet vs Opus) so you’re not burning money blindly<br> ∙ Only 6% of Claude’s 200k context window for a typical project<br> ∙ Compact mode strips comments and saves ~20% tokens automatically<br> Also has project notes that persist between sessions — your stack, architecture decisions, whatever — so Claude always has context without you typing it again.<br> The meta moment: hit my context limit building the extension itself, used it to bundle its own source, pasted into a new chat and kept going. Didn’t lose a single thing.<br> Free to install — search SendToAI in VS Code or here:<br> <a href="https://marketplace.visualstudio.com/items?itemName=OxainZ.sendtoai" rel="noopener noreferrer">https://marketplace.visualstudio.com/items?itemName=OxainZ.sendtoai</a><br> GitHub:<br> <a href="https://github.com/OxainZ/sendtoai-vscode" rel="noopener noreferrer">https://github.com/OxainZ/sendtoai-vscode</a><br> Would love feedback from anyone who tries it.​​​​​​​​​​​​​​​​</p>