AI Coding Live News

X discussion: Stripe CEO Patrick Collison adds an agent-powered "improvements" box to his vibecoded apps and calls it liquid software

A widely discussed idea on X: on Oct 31 Stripe CEO Patrick Collison (@patrickc) posted that he is building an "improvements" feature into all of his personal vibecoded applications, a freeform text input hooked up to an agent that implements whatever is submitted. He says another person who uses one of the apps has already implemented 150 improvements in a few days, and he floats the name "liquid software" for the pattern. He is also candid that he does not know whether it scales: whether it becomes unmaintainable, how collaborative software works this way, and whether it can be secure and reliable. The post drew about 2,130 likes, 1,900 bookmarks and 147K views. This is a write-up of a personal opinion and the discussion around it, not an official Stripe product announcement.

X discussion: Stripe CEO Patrick Collison adds an agent-powered "improvements" box to his vibecoded apps and calls it liquid software

Key Takeaways

  • Buzz: about 2,133 likes, 1,896 bookmarks, 217 replies, 76 quotes and 147K views (public X metrics as of Nov 1, 12:15 SGT)
  • Pattern: every personal app gets a freeform text box wired to an agent that yolo implements whatever is submitted, which makes sanding a small rough edge almost free
  • Data point: Collison says another person using one of his apps implemented 150 improvements in a few days (self-reported; the app itself is not public)
  • His own open questions: does it become unmaintainable, can model gains outpace feature appetite, how does collaborative software work, and can it be secure and reliable
  • Debate: supporters say it will become standard and could even ship bespoke per-user features behind flags; skeptics note conflicts start once a second user asks for the opposite change, and maintenance and security remain unsolved; Collison himself compares it to the user-modifiable visions of Lisp machines, Smalltalk and HyperCard
Read full story
Linked Models & Agents:

Last Secure Code Benchmark: Vibe Coding Under Fire as Top Agents Produce Secure Code on Only 25.1% of Tasks

While AI coding agents and 'vibe coding' workflows synthesize an unprecedented volume of production software, code security remains highly fragile. Researchers from NUS and Monash University introduced the Last Secure Code Benchmark, evaluating 450 tasks derived from 150 real CVE vulnerabilities across 7 languages (Java, C, C++, JS, Python, PHP, Go) and 24 CWE categories (including 21 of the 2025 CWE Top 25 and 8 of the Top 10 KEVs). While frontier agents achieve 73.6% to 91.3% functional completion rates, even the strongest model (Claude Fable 5) produces code that is both functional and secure on only 25.1% of tasks, exposing acute supply-chain risks in autonomous vibe coding.

Last Secure Code Benchmark: Vibe Coding Under Fire as Top Agents Produce Secure Code on Only 25.1% of Tasks

Key Takeaways

  • First comprehensive vibe-coding security benchmark: 450 tasks based on 150 CVEs spanning 7 languages, covering 21 of the 2025 CWE Top 25 and 8 Top 10 KEVs
  • The functionality illusion: while agents pass functional specs at 73.6%–91.3%, dual functional-and-secure pass rates collapse to 16.4%–25.1%
  • Vulnerability defense crisis: across full-repository and file generation scenarios, top models resolve only 12% of vulnerability environments securely
Read full story

gap-trap open-sources repo-specific contracts and gates for AI-written code quality

@DanKornas introduced pliablepixels/gap-trap, an open-source coding-agent skill that reads a repository, writes rules that fit the codebase, and adds gates that fail a commit or CI run when a rule is broken. It covers contracts, proven-red tests, problem ratchets, and playbooks, with multi-language gates. The repository had about 171 GitHub stars at verification and uses the MIT license.

Key Takeaways

  • Turns approved paths for HTTP, logging, settings, and auth into contracts with bypass checks.
  • Proven-red CI runs new tests against old code so a test must prove it can fail; ratchets only allow known-problem counts to fall.
  • Installs as a coding-agent skill with Node/Python and shell-based gates; pliablepixels/gap-trap had about 171 stars at verification under MIT.
Read full story

Vals releases Vibe Code Bench 1-100 for sustained multi-turn app evolution

Vals released Vibe Code Bench 1-100 (VCB 1-100), which starts from a fully passing web app and asks models to apply a long sequence of dependent product requests without breaking prior behavior; the first incomplete iteration ends the run. On the published leaderboard Claude Opus 5 leads at 28.53%, with Claude Fable 5.1 at 28.00% and GPT-6 Astra at 27.64%. Unlike one-shot greenfield coding benches, VCB 1-100 targets sustained multi-turn evolution typical of vibe coding and agentic software work.

Key Takeaways

  • VCB 1-100 starts from a fully passing app and ends on the first incomplete dependent change.
  • Leaderboard: Claude Opus 5 28.53%, Claude Fable 5.1 28.00%, GPT-6 Astra 27.64%.
  • Vals notes self-testing gaps: ~1/3 of Opus 5 and Fable 5.1 tool calls are browser use; Grok 4.6 almost never opens a browser.
Read full story
ADSponsored