Daniel Stenberg said coding agents converted 185,000 lines of C code into 340,000 lines of Rust over a little more than three weeks. During that period, the curl project merged more than 300 commits, made a little over 200 bug fixes and fixed several vulnerabilities, according to Stenberg. In the following month, he reported another 200 bug fixes and said the project may also have fixed more vulnerabilities. The figures were self-reported by Stenberg in a quotation published by Simon Willison.
Open Source Says No is a directory listing open-source projects that publicly reject AI-generated contributions, including yt-dlp, Godot, QEMU and immich. Juejin AI reported that the directory reached the Hacker News front page on November 11 with 22 verified projects, policy quotations, coverage details and verification dates. The directory aggregates existing policies rather than announcing a new policy.
VulnCheck’s seven-month review of Anthropic’s Project Glasswing says Claude generated 37,966 candidate vulnerability findings, but only 217 were marked fixed and 7,349 were made public. The tracker also counted 1,654 duplicates and 419 withdrawn findings, while Anthropic reported 629 findings with upstream patches under a different validation and remediation framework. In a curl example, maintainers rejected 19 of 34 published reports as vulnerabilities and confirmed six as low severity, highlighting the burden of human review.
cnBeta, citing The New York Times, reports that AI is accelerating the discovery of software vulnerabilities beyond the capacity of many engineers and open-source maintainers to repair them. OpenAI says researchers have used its models to find more than 1,300 vulnerabilities since it began scanning open-source software in May, while Google reports using AI to find and fix 1,072 Chrome vulnerabilities in June.
Some reports are valid and serious, but others are untested or poorly generated, adding to maintainers’ workloads; curl stopped paying bug bounties in January because of the influx. Security teams are also confronting the spread of offensive models, with Anthropic restricting its Mythos model and providing it through Project Glasswing to selected defenders.
Anthropic and Theorem report that Claude generated formal proofs of security properties at the RISC-V assembly-code level for 123 open-source libraries, with the proofs checked by Rocq. The Decoder, citing Anthropic and Theorem, says the effort produced roughly 180 million lines of proof code in a few weeks and uncovered hundreds of bugs that extensive testing had missed. These scale and bug-finding figures remain vendor-reported, while TMTPost notes that the guarantees apply only to specified properties, assumptions and verification scopes.
The curl project is changing its vulnerability disclosure process so that, from October 26, reports judged low or medium risk will be filed and discussed publicly in GitHub Issues, while high-risk and critical issues continue to be handled privately through HackerOne. Maintainer Daniel Stenberg said the steady influx of reports is overloading the maintenance team, and the new policy lets outside developers help confirm issues, check fixes and write security advisories; researchers unsure of severity can still submit to HackerOne, but low or moderate cases will be redirected to GitHub and the original report closed as informational. Stenberg said more than ninety percent of reports received in recent years could quickly be judged below high risk even though checking each still consumed substantial effort, and that many recent findings came from researchers using AI models to test libcurl options, protocols and operation sequences, with some anomalies appearing only in rarely used combinations that he considers unlikely to affect users.
Speaking at the Linux Foundation's Open Source Summit, IBM chief client innovation officer for enterprise security Jamie Thomas warned of a "tsunami in vulnerability disclosures," with roughly 66,000 unique entries expected in 2026 — about a fourfold increase over seven years ago — while time-to-exploit has shrunk from days to as little as 29 minutes. At the same time, AI-generated reports that are duplicated, low-quality or outright fake are swamping maintainers: curl shut down its HackerOne bug bounty, Google suspended its open source vulnerability rewards program, and Linus Torvalds called the Linux security mailing list "almost entirely unmanageable." Thomas argued the community should keep AI-powered vulnerability discovery but use AI to filter duplicates and bogus reports, assess severity and assist remediation, with efforts such as OpenSSF reducing the burden on maintainers.
What problem it solves
Managing multiple AI service sessions and translating them into compatible API interfaces can require repetitive account switching, custom routing, and protocol-specific integration. The project addresses these tasks with a desktop application that combines account administration, protocol conversion, request scheduling, and model routing.
Key features
Account dashboard and …
What problem it solves
AI-generated code can violate the rules its developers intended, especially when those rules are described in natural language and the resulting code is not reviewed closely. Bend addresses this by representing important rules as laws and requiring mathematical proofs when the code is edited.
Key features
Proof-enforced application laws — Applications can declare rules in `L…
What problem it solves
Organizations need to move and continuously replicate data from transactional databases, event streams, and object storage into open lakehouse formats without depending on proprietary ingestion stacks. OLake Go addresses this with replication workflows, CDC, schema handling, and deployment options that avoid separate Spark, Flink, Kafka, or Debezium infrastructure.
Key featu…
What problem it solves
CI systems report that a run failed, but often do not explain why it failed or how to fix it. This leaves teams rerunning checks, accumulating flaky failures, and delaying investigation until an experienced engineer is available.
Key features
Automated failure investigation — The agent investigates failed runs by reading logs, testing hypotheses, and gathering relevant conte…
What problem it solves
NGINX provides software for serving web traffic and handling roles including load balancing, reverse proxying, API gateway functions, and content caching. Its modular architecture, text-based configuration, and process-based runtime give users multiple ways to configure and operate those capabilities.
Key features
Modular functionality — NGINX extends its core functionality …
What problem it solves
Computer-use projects need environments where agents can operate applications, run code, and interact with graphical interfaces within the same task. They also need repeatable ways to create tasks, evaluate agents, export trajectories, and generate data for training and research.
Key features
Shared desktop Spaces — Cua Spaces gives agents full desktops hosted on computers o…
What problem it solves
Dodaios provides a locally installed, open-source way to make a personal AI assistant the control center for work on your computer. Its lead assistant coordinates models, coding agents, and computer use, while chats, memory, documents, and settings remain files on the computer that runs Dodaios.
Key features
Voice, chat, and phone interaction — The lead assistant supports wa…
What problem it solves
AI agents often discover codebase structure through repeated grep, glob, and file-reading operations, rebuilding call paths and dependencies manually. This creates unnecessary tool calls and delays before they can answer questions or make changes.
Key features
Pre-indexed semantic code graph — CodeGraph builds a local knowledge graph of symbols, call edges, and dependencies …
What problem it solves
Developers using multiple AI command-line tools need a consistent way to manage providers, accounts, prompts, MCP servers, sessions, and configuration. Switching configurations manually across applications can also make it difficult to inspect usage, preserve settings, or use different providers in separate terminal sessions.
Key features
Provider management — Manage API con…
What problem it solves
Developers need a unified way to integrate ChatGPT website capabilities into applications that expect OpenAI-compatible endpoints. They also need account-pool management, proxy configuration, import workflows, and a self-hosted interface for creating and reviewing media.
Key features
OpenAI-compatible endpoints — Provides compatible routes for image generation, image editing…
What problem it solves
Building production voice agents often requires control over infrastructure, model providers, workflows, and deployment. Dograh addresses these needs with an open-source platform that supports self-hosting, bring-your-own-key integrations, and visual and agent-assisted development.
Key features
Visual voice agent workflow builder — Create workflows with start nodes, agent no…
What problem it solves
Organizations need a distributed storage layer that can serve objects, files, tables, and vectors while keeping administration and horizontal expansion straightforward. SeaweedFS addresses this by providing these capabilities on shared storage and allowing capacity to grow by adding volume servers.
Key features
Unified distributed storage — A single cluster exposes S3 object…
What problem it solves
Shell users need a searchable, durable record of commands and terminal activity across machines, while AI agent users need continuity between sessions and harnesses. Atuin addresses both needs by recording terminal and agent activity and making it available for search, resumption, and optional synchronization.
Key features
Searchable shell history — Atuin records terminal ac…
What problem it solves
Organizations need to manage newsletters and mailing lists without relying on a hosted service. listmonk addresses this with a self-contained application that uses PostgreSQL for persistent data and can be deployed as a single binary.
Key features
Self-hosted deployment — Run the application on infrastructure you control, using either Docker Compose or the packaged binary.
S…
What problem it solves
Coding and research workflows often require persistent context, reusable operating patterns, and coordinated agents across long-running sessions. Prime Agent addresses this through a programmatic environment, durable harness state, background continuity, and mechanisms for refining supplemental state from evidence.
Key features
Persistent programmatic environment — A persist…
What problem it solves
Home networks built on OpenWrt/GL.iNet routers are spread across many devices, and there is no single place to see fleet status, per-router health, connected devices, WireGuard peers, AdGuard Home statistics or alerts as they happen. Operators also need topology and health scoring without handing their network state to a third-party cloud service. NetPulse addresses this by …
What problem it solves
Most agent frameworks require code changes, reviews, and releases to modify an agent's behavior, which puts the person who understands the desired outcome behind an engineering process. Companies also need to share agents across teams while retaining control over access, spending, approvals, and recorded activity.
Key features
Browser-based agent building — Teams can configu…
What problem it solves
Monitoring global news, geopolitical activity, markets, conflicts, infrastructure, and other live signals requires bringing information from many domains into one view. The project addresses this with a unified situational awareness interface that combines curated feeds, maps, correlation, analytics, and programmatic access.
Key features
Curated global intelligence — Curated…
What problem it solves
Running and training models locally requires tools for model execution, fine-tuning, agent integration, and deployment across different model formats and hardware. The project brings these workflows into a desktop app, a web UI, and a code-based version.
Key features
Run and train models — Run and train language, diffusion, decision, embedding, audio, image, and video models…
What problem it solves
Applications often need to turn web pages and office documents into PDFs without managing Chromium, LibreOffice, or fonts. They may also need to modify, validate, secure, and deliver the resulting PDFs through a consistent HTTP API.
Key features
Multiple document conversions — Convert HTML, URLs, and Markdown with Headless Chromium. Convert office documents with LibreOffice …