{"slug": "daniel-stenberg-agentic-rust-curl-rewrites-cannot-keep-up-with-real-curl", "title": "Daniel Stenberg: Agentic Rust curl Rewrites Cannot Keep Up With Real curl", "summary": "Curl maintainer Daniel Stenberg reported in a November 2026 blog post that coding agents converted 185,000 lines of C into 340,000 lines of Rust in a few weeks, while upstream curl shipped more than 300 commits and roughly 200 bugfixes in about three weeks that the agent-generated rewrites never absorbed. Stenberg cited his 2026 scorecard finding that only 14 of 81 confirmed curl vulnerabilities (about 17.3%) might have been avoided in Rust, with 67 remaining, and noted a November \"curl in Rust\" project spent about $8,600 in agent compute. Stenberg also described curl's OSS Scanner enrollment, where a November 5 mail cited 45 vulnerabilities, maintainers confirmed five and reclassified many as bugs, an 11% confirm rate he framed as a human-triage workflow warning rather than an anti-AI verdict.", "body_md": "**While coding agents published Rust rewrites of curl, Daniel Stenberg counts more than 300 upstream commits and roughly 200 bugfixes those clones never absorbed.** In a [November 2026 post on daniel.haxx.se](https://daniel.haxx.se/blog/actual-curl-rewrites-in-rust/?ref=explainx)—quoted the same day by [Simon Willison](https://simonwillison.net/2026/Nov/18/daniel-stenberg/?ref=explainx)—the curl maintainer says agents turned **185,000 lines of C** into **340,000 lines of Rust** in a few weeks, yet **curl itself kept shipping** vulnerability and bug work agents cannot mirror from a snapshot.\n\nFor anyone using **Claude Code, Codex, or Copilot** on legacy infrastructure code, the post is a maintenance lecture, not a language war. Stenberg already documents **[Anthropic OSS Scanner findings on curl](https://explainx.ai/blog/github-microsoft-mdash-9791-ai-findings-99-reports-november-2026)** in parallel with **[GitHub's MDASH maintainer reports](https://explainx.ai/blog/github-microsoft-mdash-9791-ai-findings-99-reports-november-2026)**; the Rust rewrite story asks a different question: **can agents operate a multi-decade ABI-stable library**, not just translate it once?\n\n## What Stenberg measured\n\nOpen box with flame linking to chips, evoking agent-generated forks that skip upstream maintenance\n\nThe core contrast is **velocity**, not syntax:\n\n| Window (per Stenberg) | curl upstream | Agent Rust rewrites | \n|---|---|---|\n| ~3 weeks | **300+ merges** ,**200+ bugfixes** ,**vulnerability fixes** | Static translation from a point-in-time tree | \n| Line count | Continuing C evolution | **185k → 340k** Rust lines claimed | \n\nStenberg names **urlx** earlier in the fall and a **November \"curl in Rust\"** project spending about **$8,600** in agent compute, per the rewrite authors' marketing. He notes AIs trained on **curl's docs, tests, and decade of issues**—translation is easier because **the specification is public**, not because agents discovered protocols.\n\n## Why one-shot rewrites miss the point\n\nStenberg's argument chains three ideas explainx.ai has seen across **[external sandboxes for coding agents](https://explainx.ai/blog/external-sandboxes-coding-agents-prime-radiant-willison-november-2026)** and **[Rust incremental rewrite tooling](https://explainx.ai/blog/rust-anything-incremental-rewrite-recorded-parity-mcp-2026)**:\n\n1. \n**curl is a time series.** Users file bugs when**real networks** misbehave. A rewrite frozen at import time**does not receive** those reports unless a team**replays** upstream history daily—something Stenberg says averages**~11 commits per day** .\n2. \n**ABI and API stability are the moat.** Only curl can show**decades** of**libcurl** compatibility. A day-old Rust tarball**might compile** ; it cannot yet prove**multi-year** integrator trust.\n3. \n**Memory safety is partial protection.** Stenberg repeats his**2026 scorecard** :**14 of 81** confirmed vulnerabilities (~**17.3%** ) might have been avoided in Rust;**67** would still exist. That aligns with**[Brotli CVE fallout across runtimes](https://explainx.ai/blog/brotli-vulnerability-cve-2026-111497-beyond-nodejs-python-bun-curl-what-to-do-2026)** —logic bugs survive language choice.\n\n## OSS Scanner versus maintainer reality\n\nBranch path reaching past a closed gate, evoking maintainer triage before AI severity labels ship\n\nStenberg's blog from **early November** walks through **OSS Scanner** enrollment after **[Max Dymond's PR](https://github.com/anthropics/oss-scanner?ref=explainx)**. The **November 5** mail cited **45 vulnerabilities**; maintainers confirmed **five**, reclassified many as **bugs**, and called **one report backwards**—patching would **introduce** the flaw the AI alleged.\n\nThat **11% vulnerability confirm rate** is not an anti-AI dunk. It is a **workflow warning**: long AI reports still need **human triage**, the same bottleneck **[Microsoft MDASH](https://explainx.ai/blog/github-microsoft-mdash-9791-ai-findings-99-reports-november-2026)** tried to ease with **private reports plus suggested fixes**.\n\n## What builders should do\n\n1. **Contribute upstream** when agents find real issues; do not fork silently.\n2. **If you migrate languages** , plan**incremental parity tests** —see[recorded parity MCP patterns](https://explainx.ai/blog/rust-anything-incremental-rewrite-recorded-parity-mcp-2026) —not a big-bang repo swap.\n3. **Track CVE closure rate** , not**LOC generated** .\n4. For **HTTP stacks in agents** , remember**[curl underpins countless tools](https://explainx.ai/blog/brotli-vulnerability-cve-2026-111497-beyond-nodejs-python-bun-curl-what-to-do-2026)** ; breaking compatibility has ecosystem blast radius.\n\n## The November rewrite wave in context\n\nStenberg is not reacting to a single repo. His **November 18** essay sits in a month where **multiple** agent-driven **curl-alikes** marketed **memory safety** and **modern Rust ergonomics** while **libcurl** consumers—**PHP**, **Git**, **Python requests stacks**, **CI pipelines**—kept filing **protocol edge cases** only visible at **Internet scale**.\n\n| Project shape (per Stenberg) | Claim | Maintainer counterpoint | \n|---|---|---|\n| **urlx** (fall 2026) | From-scratch **memory-safe** curl/libcurl | Must rebuild **trust** and**ABI** from zero | \n| **\"curl in Rust\"** (Nov 2026) | ~ **$8,600** agent spend; same branding as upstream | **Name collision** confuses integrators | \n| **curl upstream** | Still **C** | **300+ commits** and**200+ fixes** in the measured window | \n\nThe **340k Rust lines versus 185k C lines** stat is meant to puncture **\"Rust is leaner\"** marketing. More lines can mean **more generated boilerplate**, not **more capability**. Without **continuous merges**, those lines are **decorative**.\n\n## OSS Scanner: what the 45-report tarball contained\n\nStenberg's **earlier November** write-up is part of the same argument. After enrollment, **Anthropic OSS Scanner** emailed **45 alleged vulnerabilities** and **23 ordinary bugs** in one tarball. Maintainers translated each into **GitHub issues** under their **ASSESS → severity** workflow.\n\nOutcomes he reported:\n\n- **1 MEDIUM → LOW** after review\n- **44 LOW allegations → mostly bugs or NONE**\n- **5 confirmed vulnerabilities (11.1%)**\n- **1 completely inverted report** that would**create** a flaw if applied blindly\n\nHe also complains about **report length**—a familiar AI-writing pattern—and notes **11 issues still open** at the time of writing. That is the **same maintainer labor** problem **[Harvard's code-review bottleneck research](https://explainx.ai/blog/higher-scoring-coding-models-swe-bench-merge-rate-natural-experiment-2026)** describes for human teams, amplified when **every alert looks like CVE theater**.\n\nPositive read: **OSS Scanner found real flaws** curl might have missed without another pair of eyes. Negative read: **severity inflation** burns **senior maintainer weeks**. **[GitHub MDASH](https://explainx.ai/blog/github-microsoft-mdash-9791-ai-findings-99-reports-november-2026)** tried the opposite design—**private maintainer reports with patches**—because **raw alert floods** do not merge themselves.\n\n## Memory safety math Stenberg wants you to repeat\n\nStenberg's **17.3%** figure is worth quoting precisely because **CTOs use it wrongly** in **rewrite business cases**:\n\n- **81** curl vulnerabilities**confirmed in 2026** through his November tally\n- **14** plausibly**memory-safety class** in C\n- **67** would remain in a**hypothetical Rust curl**\n\nLogic bugs—**parsing mistakes**, **state-machine errors**, **protocol misunderstandings**—dominate. Agents excel at **translating syntax**; they do not automatically preserve **thirty years of negotiated behavior** around **TLS fingerprints**, **HTTP/2 push**, **FTP weirdness**, or **proxy auth** chains.\n\nIf your platform mandates **Rust for new code**, Stenberg's point is still: **refactor hot paths**, **fuzz continuously**, and **keep one canonical library**—do not ask users to relink against **unmaintained forks**.\n\n## Curl's moat is social, not syntactic\n\nStenberg's **30-year anniversary** posts (linked from the same blog index) emphasize **community velocity**:\n\n- **Hundreds of test cases** and**thousands of apps** depending on**stable ABIs**\n- **~11 commits/day** long-run average\n- **Bug reports from real users** when**CDNs change headers** or**corporate proxies break CONNECT**\n\nAgent rewrites inherit **none of that feedback loop** unless operators **pay** for **permanent merge staff**. That is why he ends **amused but puzzled**: flattery through imitation does not pay **maintainer salaries**.\n\n## Agent workflow lessons for HTTP clients in 2026\n\nTeams wiring **coding agents** into **network stacks** should translate Stenberg into policy:\n\n| Policy | Rationale | \n|---|---|\n| **Ban silent libcurl swaps** | ABI breaks **PHP, Git, CI** without loud versioning | \n| **Require upstream issue links** for agent patches | Prevents **drive-by** PRs that ignore**review backlog** | \n| **Cap rewrite spend** without**merge plan** | $8k of tokens ≠ **five years** of maintainer time | \n| **Run parity suites** on**every protocol** you actually use | Translation passes **unit tests** , fails**integration fog** | \n\nPair with **[Phonebox desk agents](https://explainx.ai/blog/phonebox-android-desk-panel-ai-coding-control-deck-2026)** and **[Soduwork multi-harness workspaces](https://explainx.ai/blog/soduwork-self-hosted-multi-harness-agent-workspace-2026)** only after **network libraries** have **owners**, not **forks**.\n\n## Thirty-year lineage in one paragraph\n\nStenberg's blog index still tells **httpget → urlget → curl** history: **Rafael Sagula's 1996 httpget**, **Daniel's maintainer takeover**, **gopher and FTP additions**, **libcurl in 2000**, **PHP adoption**, **git's use of libcurl** for remotes. Agent rewrites skip that **institutional memory**. Readers learning **why curl wins** should start with **time and trust**, not **language fashion**.\n\nWhen **[Simon Willison](https://explainx.ai/blog/external-sandboxes-coding-agents-prime-radiant-willison-november-2026)** relayed the quote, his audience—**agent builders**—got the memo: **translation demos are not operations**.\n\n## Search answers: should you replace libcurl with a Rust rewrite?\n\n**No for almost all production teams today.** If you depend on **distro packages**, **language bindings**, or **decade-old ABI promises**, stay on **upstream libcurl** and send **patches**. Consider **Rust microservices** only where you **own the full lifecycle** and can **fund merges forever**.\n\n**When might a rewrite be rational?** Greenfield **closed binaries** with **no third-party libcurl consumers**, **no need for PHP/Git compatibility**, and a **budget line for permanent parity CI**—rare in practice.\n\n**Does Anthropic OSS Scanner replace human review?** Stenberg's numbers say **no**: interesting signal, **low confirm rate on CVE labels**, heavy **triage**. Use it like **[MDASH private reports](https://explainx.ai/blog/github-microsoft-mdash-9791-ai-findings-99-reports-november-2026)**—a **feeder**, not an **owner**.\n\n**What should security teams tell developers?** Block **shadow curl forks** in **dependency scanners**; require **CVE tracking against curl.se advisories**, not **GitHub fork stars**.\n\n**Will Rust eliminate supply-chain risk?** No—**[typosquat on PyPI](https://explainx.ai/blog/microsoft-agent-framework-pypi-1-26-credential-stealer-november-2026)**-style attacks target **package names**, not **languages**. Pin **official curl** builds and teach agents **not** to suggest **alternate curl binaries** from random repos.\n\n**Bottom line:** Stenberg is not anti-Rust. He is anti-**fiction** that a **November tarball** replaces **a living project**. Measure agents on **issues closed upstream**, not **lines emitted sideways**.\n\n## Related reading\n\nReading gets you the idea. Building it with feedback gets you the skill.\n\n*Quotes and counts follow Daniel Stenberg's November 18, 2026 daniel.haxx.se post and Simon Willison's relay.*", "url": "https://wpnews.pro/news/daniel-stenberg-agentic-rust-curl-rewrites-cannot-keep-up-with-real-curl", "canonical_source": "https://explainx.ai/blog/daniel-stenberg-curl-rust-agent-rewrites-vs-maintenance-november-2026?utm_source=rss&utm_medium=feed&utm_campaign=blog", "published_at": "2026-11-18 00:00:00+00:00", "updated_at": "2026-11-18 17:16:31.959544+00:00", "lang": "en", "topics": ["ai-agents", "ai-safety", "ai-tools", "artificial-intelligence"], "entities": ["Daniel Stenberg", "curl", "Rust", "Simon Willison", "Anthropic OSS Scanner", "Max Dymond", "GitHub MDASH", "urlx"], "also_reported_by": [], "takeaway": "Agent-generated rewrites of a long-lived library like curl must replay upstream history (about 11 commits per day) rather than freeze at a snapshot, and Rust would have prevented only 14 of 81 confirmed curl vulnerabilities (~17.3%), leaving 67 logic bugs intact.", "recommendation": {"advice": "When agents rewrite or port a maintained library, wire them to replay upstream commits and bugfixes continuously instead of one-shot translation, and route AI-generated vulnerability reports through human triage before patching.", "playbook": "coding-agents", "playbook_url": "https://wpnews.pro/playbooks/coding-agents"}, "alternates": {"html": "https://wpnews.pro/news/daniel-stenberg-agentic-rust-curl-rewrites-cannot-keep-up-with-real-curl", "markdown": "https://wpnews.pro/news/daniel-stenberg-agentic-rust-curl-rewrites-cannot-keep-up-with-real-curl.md", "text": "https://wpnews.pro/news/daniel-stenberg-agentic-rust-curl-rewrites-cannot-keep-up-with-real-curl.txt", "jsonld": "https://wpnews.pro/news/daniel-stenberg-agentic-rust-curl-rewrites-cannot-keep-up-with-real-curl.jsonld"}}