<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>BuilderWithin</title><description>A practical intelligence publication for people building software in the AI era.</description><link>https://builderwithin.com/</link><language>en</language><managingEditor>hello@builderwithin.com (BuilderWithin)</managingEditor><webMaster>hello@builderwithin.com (BuilderWithin)</webMaster><item><title>A fake Claude Code 2.1.313 extension in Cursor&apos;s marketplace tried to install an SSH backdoor</title><link>https://builderwithin.com/articles/cursor-malicious-claude-code-extension-ssh-backdoor/</link><guid isPermaLink="true">https://builderwithin.com/articles/cursor-malicious-claude-code-extension-ssh-backdoor/</guid><description>Cursor says a malicious Claude Code 2.1.313 was live on October 29 and reached at most 5,000 users. Search for the svc-telemetry key to check.</description><pubDate>Fri, 30 Oct 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item><item><title>An Anthropic manager drops coding interviews and looks for prototypes</title><link>https://builderwithin.com/articles/rieseberg-anthropic-drops-coding-interviews-prototypes/</link><guid isPermaLink="true">https://builderwithin.com/articles/rieseberg-anthropic-drops-coding-interviews-prototypes/</guid><description>Felix Rieseberg has stopped giving coding interviews at Anthropic. He looks for prototypes, but says his process for assessing AI builders is unfinished.</description><pubDate>Fri, 30 Oct 2026 00:00:00 GMT</pubDate><category>signal-watch</category></item><item><title>Shared Copilot Automations run as Copilot and bill the repository owner, not the person who built them</title><link>https://builderwithin.com/articles/copilot-shared-automations-billing-identity/</link><guid isPermaLink="true">https://builderwithin.com/articles/copilot-shared-automations-billing-identity/</guid><description>GitHub&apos;s shared Copilot Automations are visible to anyone with repository read access, editable with write access, and billed to the repo owner.</description><pubDate>Thu, 29 Oct 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item><item><title>Vercel AI Gateway Virtual Models let you swap the model behind your app without redeploying</title><link>https://builderwithin.com/articles/vercel-ai-gateway-virtual-models/</link><guid isPermaLink="true">https://builderwithin.com/articles/vercel-ai-gateway-virtual-models/</guid><description>Vercel&apos;s Virtual Models give each task one model ID, so you can change the model, fallbacks and data retention in AI Gateway without code changes.</description><pubDate>Thu, 29 Oct 2026 00:00:00 GMT</pubDate><category>infra</category></item><item><title>GitHub Copilot&apos;s assisted permissions let a model approve the commands, file access and URLs it judges safe</title><link>https://builderwithin.com/articles/copilot-assisted-permissions-generally-available/</link><guid isPermaLink="true">https://builderwithin.com/articles/copilot-assisted-permissions-generally-available/</guid><description>GitHub Copilot&apos;s assisted permissions are generally available. A model screens each request and runs the clearly safe ones without asking you.</description><pubDate>Wed, 28 Oct 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item><item><title>Claude Opus 5.5 can make a game fun to play, according to four browser games built without reading the code</title><link>https://builderwithin.com/articles/simon-willison-opus-5-5-vibe-coded-games/</link><guid isPermaLink="true">https://builderwithin.com/articles/simon-willison-opus-5-5-vibe-coded-games/</guid><description>Simon Willison says Claude Opus 5.5 can design games that are fun, not just playable. His four browser games are one builder&apos;s evidence.</description><pubDate>Wed, 28 Oct 2026 00:00:00 GMT</pubDate><category>signal-watch</category></item><item><title>Cloudflare says 155 long-prompt responses on Workers AI&apos;s GLM-5.3 contained other customers&apos; data</title><link>https://builderwithin.com/articles/cloudflare-glm-5-3-workers-ai-data-exposure/</link><guid isPermaLink="true">https://builderwithin.com/articles/cloudflare-glm-5-3-workers-ai-data-exposure/</guid><description>A Cloudflare bug mixed other customers&apos; data into 155 GLM-5.3 responses to prompts over 681,280 tokens. It was fixed September 9.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate><category>infra</category></item><item><title>GPT-6.1 Terra halves standard API prices and reaches Copilot and AI Gateway</title><link>https://builderwithin.com/articles/gpt-6-1-terra-copilot-ai-gateway/</link><guid isPermaLink="true">https://builderwithin.com/articles/gpt-6-1-terra-copilot-ai-gateway/</guid><description>OpenAI says GPT-6.1 Terra halves standard API prices. Copilot and AI Gateway add it as GPT-5.6 Terra leaves Copilot November 16.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate><category>briefings</category></item><item><title>Kimi K3.1 Pro costs half as much as Claude Opus 5.5, with open weights promised by November 4</title><link>https://builderwithin.com/articles/kimi-k3-1-pro-open-weights-price/</link><guid isPermaLink="true">https://builderwithin.com/articles/kimi-k3-1-pro-open-weights-price/</guid><description>Kimi K3.1 Pro costs $2 per million input tokens and $10 for output, half the price of Claude Opus 5.5. Moonshot promises open weights by November 4.</description><pubDate>Mon, 26 Oct 2026 00:00:00 GMT</pubDate><category>briefings</category></item><item><title>Anthropic open-sources srt, a sandbox that can lock down MCP servers and other agent tools</title><link>https://builderwithin.com/articles/anthropic-sandbox-runtime-open-source-mcp-servers/</link><guid isPermaLink="true">https://builderwithin.com/articles/anthropic-sandbox-runtime-open-source-mcp-servers/</guid><description>Anthropic&apos;s open-source srt tool blocks network and file writes for any command or MCP server. It is a research preview, and Windows is alpha.</description><pubDate>Sun, 25 Oct 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item><item><title>Codex 0.164.0 adds saved prompt presets and masks credentials in remote commands</title><link>https://builderwithin.com/articles/codex-0-164-presets-credential-masking/</link><guid isPermaLink="true">https://builderwithin.com/articles/codex-0-164-presets-credential-masking/</guid><description>Codex 0.164.0 adds ten /presets slots, inline /model and /permissions arguments, and credential masking for supported remote Bash commands.</description><pubDate>Sat, 24 Oct 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item><item><title>Vercel Connect can give each agent its own restricted Stripe key</title><link>https://builderwithin.com/articles/vercel-connect-managed-stripe-keys-per-agent/</link><guid isPermaLink="true">https://builderwithin.com/articles/vercel-connect-managed-stripe-keys-per-agent/</guid><description>Vercel Connect&apos;s new Managed API Key option creates a restricted Stripe key per agent or task, so a reporting agent doesn&apos;t hold full account access.</description><pubDate>Sat, 24 Oct 2026 00:00:00 GMT</pubDate><category>infra</category></item><item><title>Issue 16: Python tools need security updates, Replit adds optional auto-publishing</title><link>https://builderwithin.com/articles/weekly-16/</link><guid isPermaLink="true">https://builderwithin.com/articles/weekly-16/</guid><description>Python tools need separate updates after the eMudhra incident. Replit adds optional automatic publishing, and Vercel budgets can pause live apps.</description><pubDate>Sat, 24 Oct 2026 00:00:00 GMT</pubDate><category>weekly-roundup</category><category>Infra &amp; Deployment</category><category>AI Builders</category><category>Coding Agents</category></item><item><title>eMudhra certificate compromise requires separate updates for pip and uv</title><link>https://builderwithin.com/articles/emudhra-ca-compromise-update-pip-uv-certifi/</link><guid isPermaLink="true">https://builderwithin.com/articles/emudhra-ca-compromise-update-pip-uv-certifi/</guid><description>The eMudhra certificate incident affects Python tools used by coding agents. Updating your operating system alone does not fix older pip installs.</description><pubDate>Fri, 23 Oct 2026 00:00:00 GMT</pubDate><category>infra</category></item><item><title>Replit Monitoring Agent fixes production errors and, with auto-publish on, deploys the fix itself</title><link>https://builderwithin.com/articles/replit-monitoring-agent-auto-publish-fixes/</link><guid isPermaLink="true">https://builderwithin.com/articles/replit-monitoring-agent-auto-publish-fixes/</guid><description>Replit&apos;s Monitoring Agent investigates errors in published apps and validates fixes. With auto-publish on, it deploys them to your live app itself.</description><pubDate>Fri, 23 Oct 2026 00:00:00 GMT</pubDate><category>ai-builders</category></item><item><title>Claude Mods let you control when Claude Code interrupts you, and noise is the main risk</title><link>https://builderwithin.com/articles/claude-mods-attention-interface-mcateer/</link><guid isPermaLink="true">https://builderwithin.com/articles/claude-mods-attention-interface-mcateer/</guid><description>Dan McAteer argues Claude Code mods let you shape what the agent shows you and when. His own mod fired on 99 of 111 turns, so noise is the test.</description><pubDate>Wed, 21 Oct 2026 00:00:00 GMT</pubDate><category>signal-watch</category></item><item><title>Supabase Agent prompts let your coding agent check production on a schedule</title><link>https://builderwithin.com/articles/supabase-agent-prompts-production-checks/</link><guid isPermaLink="true">https://builderwithin.com/articles/supabase-agent-prompts-production-checks/</guid><description>Supabase&apos;s Agent prompts have your coding agent check a live project hourly or daily for errors, security issues and slow queries, then report findings.</description><pubDate>Wed, 21 Oct 2026 00:00:00 GMT</pubDate><category>infra</category></item><item><title>Cloudflare open-sources Void, which deploys Vite apps and provisions their resources</title><link>https://builderwithin.com/articles/cloudflare-void-vite-full-stack-deploy/</link><guid isPermaLink="true">https://builderwithin.com/articles/cloudflare-void-vite-full-stack-deploy/</guid><description>Cloudflare&apos;s open-source Void toolkit deploys Vite apps to Cloudflare and sets up their databases, with a shared platform for internal tools.</description><pubDate>Sat, 17 Oct 2026 00:00:00 GMT</pubDate><category>infra</category></item><item><title>OpenCode ends session sharing, and existing share links stop working</title><link>https://builderwithin.com/articles/opencode-ends-session-sharing/</link><guid isPermaLink="true">https://builderwithin.com/articles/opencode-ends-session-sharing/</guid><description>OpenCode 1.19.1 discontinued session sharing. Old share links and sharing commands no longer work, and sessions can&apos;t be imported from share URLs.</description><pubDate>Sat, 17 Oct 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item><item><title>Issue 15: Gemini 4 Argon&apos;s API price doubles after November 30</title><link>https://builderwithin.com/articles/weekly-15/</link><guid isPermaLink="true">https://builderwithin.com/articles/weekly-15/</guid><description>Gemini 4 Argon&apos;s API price doubles after November 30, and Copilot retires three models on November 16. Cursor&apos;s Grok plan bills other models extra.</description><pubDate>Sat, 17 Oct 2026 00:00:00 GMT</pubDate><category>weekly-roundup</category><category>Coding Agents</category><category>AI Builders</category><category>Infra &amp; Deployment</category></item><item><title>GitHub Copilot retires Claude Opus 5, Opus 4.8 and GPT-5.6 Terra on November 16</title><link>https://builderwithin.com/articles/github-copilot-opus-gpt-deprecations-november-16/</link><guid isPermaLink="true">https://builderwithin.com/articles/github-copilot-opus-gpt-deprecations-november-16/</guid><description>GitHub Copilot removes Claude Opus 5, Opus 4.8 and GPT-5.6 Terra on November 16. Annual individual plans keep Opus 4.8; others should switch.</description><pubDate>Fri, 16 Oct 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item><item><title>Lovable now plans and drafts ad and SEO campaigns from the same chat that builds your app</title><link>https://builderwithin.com/articles/lovable-growth-tools-ads-seo-campaigns/</link><guid isPermaLink="true">https://builderwithin.com/articles/lovable-growth-tools-ads-seo-campaigns/</guid><description>Lovable can build a growth plan and ad assets for your app. Ad spend bills through Google or Meta, and you approve budget before launch.</description><pubDate>Fri, 16 Oct 2026 00:00:00 GMT</pubDate><category>ai-builders</category></item><item><title>Cheap clones are squeezing software revenue, and Peter Yang says agent access is where money moves</title><link>https://builderwithin.com/articles/peter-yang-software-built-for-agents-first/</link><guid isPermaLink="true">https://builderwithin.com/articles/peter-yang-software-built-for-agents-first/</guid><description>Peter Yang says traditional software is harder to monetize as AI clones it, and that money is moving to products agents can use directly.</description><pubDate>Fri, 16 Oct 2026 00:00:00 GMT</pubDate><category>signal-watch</category></item><item><title>Copilot adds HydraFusion Max and increases reasoning effort for review</title><link>https://builderwithin.com/articles/github-copilot-hydrafusion-max-review-billing/</link><guid isPermaLink="true">https://builderwithin.com/articles/github-copilot-hydrafusion-max-review-billing/</guid><description>Copilot&apos;s HydraFusion adds a Max option and deeper review. Each model pass is billed, and discarded drafts can leave file edits behind.</description><pubDate>Thu, 15 Oct 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item><item><title>Vercel adds free Nemotron 3.5 Lightning, with a prompt-retention warning</title><link>https://builderwithin.com/articles/vercel-nemotron-lightning-free-prompt-retention/</link><guid isPermaLink="true">https://builderwithin.com/articles/vercel-nemotron-lightning-free-prompt-retention/</guid><description>Vercel&apos;s free Nemotron 3.5 Lightning route is served by NVIDIA, which may retain prompts for model improvement. Check what you send.</description><pubDate>Thu, 15 Oct 2026 00:00:00 GMT</pubDate><category>infra</category></item><item><title>Muse Wonder ranks first on three agent tasks but produces weak design, Claire Vo&apos;s benchmark finds</title><link>https://builderwithin.com/articles/claire-vo-muse-wonder-agents-not-design/</link><guid isPermaLink="true">https://builderwithin.com/articles/claire-vo-muse-wonder-agents-not-design/</guid><description>Claire Vo&apos;s roughly 20-task benchmark found Meta&apos;s Muse Wonder ranked first on three agent tasks, produced weak design, and ran slowly on agent work.</description><pubDate>Wed, 14 Oct 2026 00:00:00 GMT</pubDate><category>signal-watch</category></item><item><title>Claude Code now summarizes long conversations you leave idle for about 54 minutes</title><link>https://builderwithin.com/articles/claude-code-idle-compaction-prompt-cache/</link><guid isPermaLink="true">https://builderwithin.com/articles/claude-code-idle-compaction-prompt-cache/</guid><description>Claude Code 2.1.299 compacts a 200K+ token terminal conversation after about 54 idle minutes. You can raise the threshold or turn it off.</description><pubDate>Wed, 14 Oct 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item><item><title>Google launches Gemini 4 Argon for complex coding tasks</title><link>https://builderwithin.com/articles/gemini-4-argon/</link><guid isPermaLink="true">https://builderwithin.com/articles/gemini-4-argon/</guid><description>Gemini 4 Argon reaches coding agents with introductory API prices that double after November 30. Access and billing depend on the tool you use.</description><pubDate>Wed, 14 Oct 2026 00:00:00 GMT</pubDate><category>briefings</category></item><item><title>Cursor joins the X Subscription with shared Grok usage</title><link>https://builderwithin.com/articles/cursor-x-subscription-grok-usage-pool/</link><guid isPermaLink="true">https://builderwithin.com/articles/cursor-x-subscription-grok-usage-pool/</guid><description>Cursor joins xAI&apos;s X Subscription. Grok usage shares one weekly pool across apps; other models in Cursor cost extra at provider rates.</description><pubDate>Tue, 13 Oct 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item><item><title>Lovable adds EU inference for Enterprise workspaces</title><link>https://builderwithin.com/articles/lovable-eu-inference-enterprise/</link><guid isPermaLink="true">https://builderwithin.com/articles/lovable-eu-inference-enterprise/</guid><description>Lovable&apos;s Enterprise EU inference setting routes supported AI requests to European endpoints and fails rather than sending them elsewhere.</description><pubDate>Tue, 13 Oct 2026 00:00:00 GMT</pubDate><category>ai-builders</category></item><item><title>Vercel CLI lets agents find domains and hand purchases back for approval</title><link>https://builderwithin.com/articles/vercel-cli-agents-buy-domains-confirmation/</link><guid isPermaLink="true">https://builderwithin.com/articles/vercel-cli-agents-buy-domains-confirmation/</guid><description>Vercel&apos;s CLI lets a coding agent search domains and check prices, but buying hands back to you for the price, auto-renew, and contact details.</description><pubDate>Sun, 11 Oct 2026 00:00:00 GMT</pubDate><category>infra</category></item><item><title>Vercel Pro Teams Have Until October 23 to Keep Older Deployments</title><link>https://builderwithin.com/articles/vercel-pro-30-day-deployment-retention/</link><guid isPermaLink="true">https://builderwithin.com/articles/vercel-pro-30-day-deployment-retention/</guid><description>Vercel extends 30-day deployment retention to existing Pro teams. Opt out before October 23 to keep older deployments available for rollback.</description><pubDate>Sat, 10 Oct 2026 00:00:00 GMT</pubDate><category>infra</category></item><item><title>Issue 14: Broken Claude Code Hooks Can Stop Work, Vercel Sets a Rollback Deadline</title><link>https://builderwithin.com/articles/weekly-14/</link><guid isPermaLink="true">https://builderwithin.com/articles/weekly-14/</guid><description>Claude Code hooks can block on failure when configured. Vercel Pro teams must opt out before October 23 to keep older deployments.</description><pubDate>Sat, 10 Oct 2026 00:00:00 GMT</pubDate><category>weekly-roundup</category><category>Coding Agents</category><category>Infra &amp; Deployment</category><category>AI Builders</category><category>Workflow</category></item><item><title>Claude Code can block tool calls when a guardrail hook fails</title><link>https://builderwithin.com/articles/claude-code-hooks-onfailure-block/</link><guid isPermaLink="true">https://builderwithin.com/articles/claude-code-hooks-onfailure-block/</guid><description>Claude Code 2.1.295 adds onFailure &quot;block&quot; for command and HTTP hooks. Before-action hooks can stop tool calls when a check fails.</description><pubDate>Fri, 09 Oct 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item><item><title>OpenAI&apos;s Ultrafast Tier on Vercel AI Gateway Costs 6x the Standard Rate</title><link>https://builderwithin.com/articles/vercel-ai-gateway-openai-ultrafast-6x-price/</link><guid isPermaLink="true">https://builderwithin.com/articles/vercel-ai-gateway-openai-ultrafast-6x-price/</guid><description>Vercel AI Gateway now offers OpenAI&apos;s Ultrafast tier for GPT 6 Astra and GPT 6.1 Sol at 6x per-token pricing, with EU-pinned requests falling back.</description><pubDate>Fri, 09 Oct 2026 00:00:00 GMT</pubDate><category>infra</category></item><item><title>Claude Haiku 5.5 brings lower-cost helper tasks to coding agents</title><link>https://builderwithin.com/articles/claude-haiku-5-5-subagent-model-copilot-gateway/</link><guid isPermaLink="true">https://builderwithin.com/articles/claude-haiku-5-5-subagent-model-copilot-gateway/</guid><description>Claude Haiku 5.5 lowers helper-task costs in Claude Code, Copilot, and AI Gateway. Prompts over 100,000 tokens use a higher rate.</description><pubDate>Thu, 08 Oct 2026 00:00:00 GMT</pubDate><category>briefings</category></item><item><title>Cursor&apos;s iOS app can message local agents while your computer stays awake</title><link>https://builderwithin.com/articles/cursor-remote-control-local-agents-ios/</link><guid isPermaLink="true">https://builderwithin.com/articles/cursor-remote-control-local-agents-ios/</guid><description>Cursor&apos;s iOS app can now see and reply to local agents on your computer. It is on by default outside Enterprise, and your machine has to stay awake.</description><pubDate>Wed, 07 Oct 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item><item><title>Nathan Lambert argues cyber-risk restrictions must account for closed APIs too</title><link>https://builderwithin.com/articles/open-weight-cyber-risk-ban-closed-apis-lambert/</link><guid isPermaLink="true">https://builderwithin.com/articles/open-weight-cyber-risk-ban-closed-apis-lambert/</guid><description>Nathan Lambert argues open-model restrictions must account for hosted APIs, defensive uses, and gaps in publicly documented cyber incidents.</description><pubDate>Wed, 07 Oct 2026 00:00:00 GMT</pubDate><category>signal-watch</category></item><item><title>v0 moves personal accounts into team workspaces</title><link>https://builderwithin.com/articles/v0-personal-accounts-team-workspaces-chat-visibility/</link><guid isPermaLink="true">https://builderwithin.com/articles/v0-personal-accounts-team-workspaces-chat-visibility/</guid><description>v0 moved personal accounts into team workspaces. Your chats came along, and new chats in a team workspace are visible to the team by default.</description><pubDate>Wed, 07 Oct 2026 00:00:00 GMT</pubDate><category>ai-builders</category></item><item><title>Claude Code keeps local session transcripts for 30 days by default</title><link>https://builderwithin.com/articles/claude-code-session-transcripts-30-day-cleanup/</link><guid isPermaLink="true">https://builderwithin.com/articles/claude-code-session-transcripts-30-day-cleanup/</guid><description>Claude Code’s cleanupPeriodDays setting controls transcript retention, and longer retention leaves plaintext session history on disk.</description><pubDate>Mon, 05 Oct 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item><item><title>Simon Willison calls for default spending caps on agent-run services</title><link>https://builderwithin.com/articles/willison-default-hard-budget-caps-agents/</link><guid isPermaLink="true">https://builderwithin.com/articles/willison-default-hard-budget-caps-agents/</guid><description>Simon Willison says pay-by-usage services need hard budget caps on by default. AWS and Google Cloud now offer them, with different shutoff rules.</description><pubDate>Mon, 05 Oct 2026 00:00:00 GMT</pubDate><category>signal-watch</category></item><item><title>Supabase brings backend settings into project files for coding agents</title><link>https://builderwithin.com/articles/supabase-schema-config-in-repo-agents/</link><guid isPermaLink="true">https://builderwithin.com/articles/supabase-schema-config-in-repo-agents/</guid><description>The declarative schema flow is experimental, local testing is alpha, and Supabase Compute remains in private alpha with a waitlist.</description><pubDate>Sun, 04 Oct 2026 00:00:00 GMT</pubDate><category>infra</category></item><item><title>Supabase adds agent confirmations that depend on MCP client support</title><link>https://builderwithin.com/articles/supabase-select-agent-permissions-elicitations/</link><guid isPermaLink="true">https://builderwithin.com/articles/supabase-select-agent-permissions-elicitations/</guid><description>Supabase&apos;s MCP server asks you to confirm paid resources and data-deleting SQL, but prompts appear only in agent clients that support MCP elicitations.</description><pubDate>Sat, 03 Oct 2026 00:00:00 GMT</pubDate><category>infra</category></item><item><title>Issue 13: Agent access expands from code to desktops and databases</title><link>https://builderwithin.com/articles/weekly-13/</link><guid isPermaLink="true">https://builderwithin.com/articles/weekly-13/</guid><description>Claude Code mods can change permission handling. Supabase adds agent confirmations, and Lovable&apos;s security fix depends on where your app runs.</description><pubDate>Sat, 03 Oct 2026 00:00:00 GMT</pubDate><category>weekly-roundup</category><category>Coding Agents</category><category>Infra &amp; Deployment</category><category>AI Builders</category><category>Workflow</category></item><item><title>Claude Code mods run unsandboxed and can alter permission checks</title><link>https://builderwithin.com/articles/claude-code-mods-unsandboxed-plugins/</link><guid isPermaLink="true">https://builderwithin.com/articles/claude-code-mods-unsandboxed-plugins/</guid><description>Claude Code mods run with access to your machine, so an installed plugin can change the safety and permission checks you rely on.</description><pubDate>Fri, 02 Oct 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item><item><title>Hamel Husain says Claude&apos;s eval workflow needs data review before test selection</title><link>https://builderwithin.com/articles/hamel-husain-claude-auto-evals-look-at-data/</link><guid isPermaLink="true">https://builderwithin.com/articles/hamel-husain-claude-auto-evals-look-at-data/</guid><description>Hamel Husain tried Anthropic&apos;s new Claude Code eval tool and says it skips the step that matters most, which is reading your own data first.</description><pubDate>Fri, 02 Oct 2026 00:00:00 GMT</pubDate><category>signal-watch</category></item><item><title>Cloudflare Workers sends grouped production errors to coding agents</title><link>https://builderwithin.com/articles/cloudflare-workers-issues-error-monitoring-agents/</link><guid isPermaLink="true">https://builderwithin.com/articles/cloudflare-workers-issues-error-monitoring-agents/</guid><description>Cloudflare&apos;s new Issues feature groups repeated Workers errors and hands the stack trace and logs to Claude Code, Cursor, or Devin. It&apos;s in open beta.</description><pubDate>Thu, 01 Oct 2026 00:00:00 GMT</pubDate><category>infra</category></item><item><title>GPT-6.1 Sol keeps Sol&apos;s API rates with cheaper cached input</title><link>https://builderwithin.com/articles/gpt-6-1-sol-coding-price-tradeoff/</link><guid isPermaLink="true">https://builderwithin.com/articles/gpt-6-1-sol-coding-price-tradeoff/</guid><description>GPT-6.1 Sol keeps GPT-6 Sol&apos;s standard API rates, halves cached input costs, and rejects two lower reasoning settings GPT-6 Sol supports.</description><pubDate>Thu, 01 Oct 2026 00:00:00 GMT</pubDate><category>briefings</category></item><item><title>Lovable blocks TanStack Start exploit traffic; exported apps need the patch</title><link>https://builderwithin.com/articles/lovable-tanstack-start-xss-vulnerability/</link><guid isPermaLink="true">https://builderwithin.com/articles/lovable-tanstack-start-xss-vulnerability/</guid><description>Lovable added firewall protection for a TanStack Start flaw. Affected apps hosted elsewhere need the framework update and redeployment.</description><pubDate>Thu, 01 Oct 2026 00:00:00 GMT</pubDate><category>ai-builders</category></item><item><title>Codex Pro 200 reopens with lower usage; eligible subscribers keep their allowance until October 29</title><link>https://builderwithin.com/articles/codex-plan-multipliers-pro-200-reopens/</link><guid isPermaLink="true">https://builderwithin.com/articles/codex-plan-multipliers-pro-200-reopens/</guid><description>New Pro 200 subscriptions have lower included usage. Eligible existing subscribers keep their previous allowance through October 29.</description><pubDate>Wed, 30 Sep 2026 00:00:00 GMT</pubDate><category>coding-agents</category></item></channel></rss>