Skip to content
#Deployment / engineering
November 13Friday
  1. Hacker News · AI38

    Show HN: Build Django Admin internal tools with your coding agent

    Django Control Room (DCR) lets a coding agent scaffold internal tools inside Django Admin, starting with a `/plugin install dcr --marketplace django-control-room/skills` command. The hub collects every tool a team builds, whether in-project or shipped as a package, and exposes each panel's tools to AI agents through one endpoint under the same permissions. Panels are staff-only by default and pick up Unfold, Jazzmin, Grappelli and other admin themes automatically.

  2. Hacker News · AI71

    Kiln-AI releases Harness Tuner, an eval-driven tuning loop for agent tools, prompts and models

    Kiln-AI released Harness Tuner, an open-source CLI that tunes an agent's tools, prompts, skills, model and settings against the evals you already have. On one production multi-turn agent with over 200 tools, the team reports 2.5x fewer failures, 6x lower cost and 2x lower latency on held-out data with a smaller model. It runs on Claude Code, Codex or OpenCode subscriptions, keeps train, val and test splits enforced in code, and works with the free Kiln app for experiment tracking.

  3. Hacker News · AI58

    Skylands launches to run coding agents on any machine and approve them from a phone

    Skylands is a new tool that gives developers a single browser interface for dispatching coding agent work to any machine they can access, and for approving permission requests from a phone. It supports Claude Code, Gemini, Codex and Goose, and keeps a secure signed log of agent actions and approvals plus a central inbox for conversation history. It installs via Homebrew on macOS or Linux, and `skylands devworld up` starts an embedded NATS server, a citadel and a daemon in one process.

  4. Hacker News · AI55

    Langfuse releases AI Gateway, an open source Rust gateway that traces every LLM call

    Langfuse has published AI Gateway, a standalone Rust service that proxies native OpenAI and Anthropic API paths under /openai/v1 and /anthropic/v1 and uploads each finalized call to Langfuse as a generation over OTLP. The gateway relays provider JSON and SSE bytes without rewriting them, resolves credentials through Langfuse Web, and works with Claude Code by pointing ANTHROPIC_BASE_URL at the /anthropic namespace. It ships as a multi-platform Docker image on Docker Hub and ghcr.io, with capture limited to 5 MiB request inspection and 1 MiB retained output per execution.

  5. Hacker News · AI66

    Devin launches Apps, private web apps built and hosted on its cloud machine

    Devin launched Devin Apps, private web applications that Devin builds and hosts directly on its cloud computer, reachable from a stable URL that requires Devin sign-in. Apps can use existing MCPs, integrations and codebase, start from templates such as a metrics dashboard or review queue, and be updated by asking Devin to edit in session, with each change kept as a version under Deployments. The machine sleeps when idle and billing is based on awake time, and each App runs on one machine, which the company positions for internal tools and dashboards rather than high-traffic production services.

  6. Hacker News · AI58

    DeepNoodle releases Decide, a decision-model CLI for auditing large codebases

    DeepNoodle released Decide, an open-source Apache 2.0 CLI that sends one question per item to a decision model and returns a calibrated probability instead of generated text. It splits Go, Python, JavaScript, TypeScript and Java into functions, caches every answer so reruns only ask about what changed, and outputs JSON, CSV, Markdown or GitHub annotations. Two Claude Code plugin skills use it to score functions for security bugs or repeated mistakes, and Claude then reads only the top 15. Jev is the default model at $0.042 per million input tokens, with TypeSafe, Cloudflare Clef, OpenAI Luna and Google Gemini selectable.

  7. TechCrunch · AI55

    ElevenLabs releases ElevenAI preview app combining meeting notes, dictation and agents

    ElevenLabs released an experimental preview app called ElevenAI that combines meeting note-taking, dictation and agent integration, launched at its 11/11 Summit in New York with sign-up required for early access. The app records and notifies meeting participants, lets users query data across meetings, offers live in-chat assistance, and supports dictation in over 90 languages with industry-specific term recognition. Companies can also connect ElevenAgent to take follow-up actions such as drafting emails from meeting notes, as ElevenLabs positions the product for teams rather than individual productivity.

  8. Hacker News · AI45

    Deccy: A Decision Management Framework for AI-Assisted Software Engineering

    Deccy is a pre-alpha decision management framework for doing software engineering with AI, splitting human high-level decision-making from AI execution. It ships as deccy-cli, a set of deccy-agent custom agents, and deccy-viewer, which visualizes decisions and their relationships as a directed acyclic graph where every decision carries a rationale and is evaluated against its dependencies. The team targets long-term projects where intent gets lost, and reports anecdotally that coding agents need less steering on deccy-modeled projects.

  9. Hacker News · AI58

    A Deep Dive into Decision Models: How Perplexity Decider v1.2 Works

    This post dissects Perplexity Decider v1.2, an open decision model built on Qwen-3.8 27B that converts a decoder-only LLM into a non-causal encoder and answers all classification questions in a single prefill step via a 255-option readout head. It covers the /v1/decisions API's yes_no, choice and score question types, training on harmonized datasets such as tasksource with synthetic data and choice shuffling, and post-hoc calibration at temperature T=1.32. The author argues the main case for decision models over autoregressive LLMs is token efficiency, and notes the 9B Decider Flash variant is trained with a distillation loss.

  10. Hacker News · AI22

    How Microsoft Principal Engineering Manager Tarik Guney Uses AI to Remove Bottlenecks

    Tarik Guney, a Principal Software Engineering Manager on Microsoft Teams, treats AI as a tool that extends engineers' physical and cognitive limits rather than a capability to chase. He focuses on leading indicators and bottlenecks across culture, goals, and tooling, arguing that process changes alone raise cognitive load instead of reducing it.

  11. Hacker News · AI58

    Airdash launches SQLite databases for AI agents with an Airtable-style view

    Airdash launched a product that lets Claude, Codex and other MCP clients create and maintain Cloudflare D1 SQLite databases through an MCP server at /mcp, while users read and edit the data in a spreadsheet view. Each database can be rolled back to any minute in the last 30 days, downloaded as a standard SQL file, and shared with teammates; the free plan offers 100 databases, 1 GB of storage and 1,000,000 rows written per month.

  12. Hacker News · AI40

    Let an AI agent query your production database safely with read-only SQL over MCP

    This tutorial shows how to let an AI agent answer plain-language questions about a production database on Railyard using the list_tables and query_database MCP tools. Every query passes five guard rails: write-scoped connections, admin/owner-only role caps, read-only transactions that always roll back, a statement timeout, and a 200-row cap flagged as truncated. The SQL text is never written into the MCP audit entry.

  13. Hugging Face Blog58

    llama.cpp Releases v1.0.0, Stabilizing Downstream APIs with Semantic Versioning Rules

    llama.cpp has released v1.0.0, and ggml has also released v1.0.0. Going forward, both projects will indicate incompatible changes to the public C API and ABI through major version number changes. The project team states that as local AI and privacy needs grow and downstream applications increase, stable library interfaces have become critical, and the boundaries for breaking changes will be stricter in the future.

  14. Hacker News · AI38

    CloudHike: A Backend That AI Coding Agents Can Set Up via MCP

    CloudHike launched a backend platform that AI coding agents can configure over MCP, letting Cursor, Claude Code, Codex and VS Code create projects and deploy services from a conversation. It bundles Postgres, auth, storage, Redis, email, jobs and 100+ AI models behind one API key, priced at $10/mo versus $25/mo for Supabase and others. A free plan includes $10 in signup credits with no credit card required.

  15. Hacker News · AI65

    Brainfish AI open-sources Webcrawl, a zero-key web crawling API and MCP server for agents

    Brainfish AI released Webcrawl, an open-source web crawling API and MCP server that lets agents such as Claude Code, Cursor and Codex search, scrape, map, crawl and monitor sites without an API key. One command, npx webcrawl setup, registers the @webcrawl/mcp server and installs the webcrawl-anything Agent Skill, while npx skills add brainfish-ai/webcrawl-anything installs all eight skills, including ticket, job, deal and real-estate watchers. The same TypeScript engine runs locally and in the hosted webcrawl.app cloud, and the project is MIT licensed.

  16. Hacker News · AI50

    NukeTorch: Four MacBook Pro M5 Max Macs Serve NVIDIA GR00T 2.45x Faster Than Fastest Published H100 Result

    Fenon AI's NukeTorch engine ran NVIDIA's GR00T N1.7 robot policy on four MacBook Pro M5 Max machines at 152.07 action chunks per second with 192 clients, 2.45x the fastest published H100 figure and 4.24x NVIDIA's own published number. For a single request the H100 still wins at 16.08 ms versus 44.34 ms, and the four Macs list at $19,996 against at least $25,000 for one H100 card. The one-Mac version is a free download, while the multi-machine build requires an evaluation license.

  17. NVIDIA Newsroom22

    How AI Is Helping Care Teams Get Patients the Right Care, Sooner

    Dyania Health's Synapsis AI clinical reasoning engine reads every note, report and lab result in a patient's record, connecting findings years apart to flag which patients need attention and why, with each signal traceable to its source. It runs on NVIDIA-accelerated computing inside a health system's own firewall, so patient data never leaves the hospital. Cleveland Clinic used it to review records of more than 7,800 aortic valve replacement patients from 2005 to 2021 and build a predictive model of valve deterioration and reintervention risk at three, five and ten years.

  18. Hacker News · AI22

    AI Room Design tool offers 2 free trial designs from a room photo

    An AI interior design tool turns a photo of your own room into restyled concepts across six styles, covering both furnished-room redesign and virtual staging for empty spaces. New users get 2 free trial designs after signing in, with trial credits expiring after 30 days, while paid one-time packs start at $9.99 for 60 designs (about $0.17 each) and purchased credits never expire. Designs are AI-generated concepts rather than measured plans, so furniture sizes and structure should be verified before building.

  19. Hacker News · AI38

    Rethinking Memory for AI Inference

    Moving compute closer to memory — via processing near memory (PNM) and processing in memory (PIM) — can cut data movement, power consumption and latency for AI inference at the edge. Cadence's Frank Ferro says LLMs moving into physical devices are forcing architects to rethink memory partitioning, while Quadric's Steve Roddy notes inference's known data access patterns suit ahead-of-time compilation. Software support, standardization and thermal management remain barriers to adoption.

  20. Hacker News · AI48

    The unlikely seaside town on frontline of AI threat

    Bournemouth's Littledown campus, a former farm site, now hosts over 5,300 JP Morgan staff processing more than $1tn in global transactions daily, making Dorset one of Britain's largest financial hubs outside London. Regional finance centres face an existential AI threat as bosses weigh automating back-office roles, after Barclays announced 5,000 AI-linked job cuts and Standard Chartered planned 7,000.

  21. MarkTechPost58

    Google releases TPU Sync, an open-source KV cache transfer library for Cloud TPUs

    Google Cloud has released TPU Sync, an Apache-2.0 C++ library that moves KV caches and model weights across TPU HBM, pinned host DRAM, shared storage and the data center network. In internal July 2026 tests on TPU v7x, Google reports over 90% lower KV transfer overhead, with Qwen3 235B median KV transfer latency dropping from several seconds to about 230 ms and average TTFT falling 60% for 235B and 90% for 480B at low concurrency. The library integrates with PyTorch via TorchTPU and with JAX, serving as the KV connector in vLLM-TorchTPU and the P/D transfer backend in SGLang-TorchTPU, though the README says it is not yet recommended for general use.

  22. Hacker News · AI36

    Agent-brain: A Notes-Agnostic Second-Brain Operating Model for AI Coding Agents

    Agent-brain is a personal operating model for AI coding agents such as Claude Code, OpenCode, Codex and Antigravity CLI, built on a second-brain structure of journal, WIP, memory and tasks that works over any folder of notes, with Obsidian optional. Runtime config and memory like CLAUDE.md and AGENTS.md are kept in a git-tracked brain and symlinked into each runtime, so setup travels across machines. Install runs via curl and dry-runs by default, with a session lifecycle driven by the brain skill.

  23. Hacker News · AI62

    How to design effective tools for AI agents: guidance from OpenAI, Anthropic, Google and MCP

    This post consolidates what OpenAI, Anthropic, Google and the MCP spec recommend for designing tools for AI agents, and covers what their guides leave out. It argues tools should be built around workflows rather than API endpoints, that names, descriptions and schemas act as a contract the model reads literally, and that responses and errors are part of the prompt.

  25. Hacker News · AI71

    Google Cloud launches Gemini for healthcare in preview for payers and providers

    Google Cloud announced Gemini for healthcare, a purpose-built industry solution that extends the Gemini agent introduced last month at Gemini at Work, and it is available in preview in the U.S. starting today. The offering combines reusable skills for clinical and administrative tasks, connectors to EHR, claims and care management systems such as Epic, Oracle Health and InterSystems FHIR, a partner ecosystem, and a governed platform built to support HIPAA compliance with audit logging and verifiable citations. Google says clinical data, prompts and outputs stay in the organization's governed environment and are never used to train or fine-tune its foundation models.

  26. Hacker News · AI58

    Anthropic says its AI model moved Claude onto AMD chips in three days

    Anthropic told investors that its Mythos model got Claude running on AMD MI355X GPUs in about three days, with Mythos writing or porting nearly all the necessary kernels and, in co-founder Tom Brown's words, basically no human intervention. The timeline covers bringing up prompt processing for one Anthropic model at a useful performance level, not a full migration of Claude's training and operations. Anthropic says its compute architecture spans Nvidia GPUs, AMD GPUs, AWS Trainium and Google TPUs, treated as a shared capacity pool.

  27. Hacker News · AI58

    Uber Open-Sources Michelangelo ML Platform, Says It Has Been Validated Across 400+ Projects and 20,000 Monthly Training Jobs

    Uber has open-sourced Michelangelo, its ML platform proven at Uber scale with 400+ active ML projects, 20K+ model training jobs per month, 5K+ models in production and 45M+ real-time predictions per second at peak. The release covers feature management, distributed model training, deployment and monitoring, with feature management, evaluation and monitoring marked coming soon. A Python API built on Uniflow lets users define distributed training tasks and end-to-end workflows, such as a Ray-based XGBoost training pipeline, in a few lines of code.

  28. Hacker News · AI38

    Show HN: FloxoLab AI Crawler Access Checker Tests 14 AI Crawlers

    A developer has launched the FloxoLab AI Crawler Access Checker, a free tool that tests whether 14 AI crawlers can actually read a given page. It reads robots.txt per RFC 9309, re-requests the URL with each crawler's user agent, and compares status codes and headers against a normal browser request, returning a verdict per crawler. The tool is free with no account and limited to 8 checks per 10 minutes.

  29. MIT News34

    MIT researchers build an AI "copilot" for topology optimization in structural design

    MIT researchers developed an AI copilot that lets engineers give feedback on regions of a structure while topology optimization is still running, instead of waiting for the final design. In demonstrations it improved manufacturability and linear buckling load by 39 percent while adding only 15 seconds to total design time. The system uses machine learning trained on synthetic human feedback data, and currently only lets users thicken or thin regions.

  30. Hacker News · AI36

    AI Security Wiki: An Implementation Reference for Enterprise AI Security

    The AI Security Wiki is a reference for securing enterprise AI adoption, covering governance, risk management, secure engineering, operational monitoring, vendor evaluation and assurance in reusable template pages. It maps control objectives, audit evidence and test cases onto frameworks including ISO/IEC 42001, NIST AI RMF, NIST SSDF, PCI DSS v4.0.1, the FINOS AI Governance Framework and CERT-In's AI security auditing guidelines. It also includes a 682-case test library drawn from OWASP lists, MITRE ATLAS and NIST AI 100-2.

  31. Hacker News · AI22

    Glue work is still harmful in the age of AI

    LLM-written code has not made glue work more valuable; it has made most of it less necessary. AI agents let one engineer handle projects that once needed a small team, cutting coordination meetings, while fewer junior hires and LLM-generated write-only reports remove other glue tasks. The author still warns engineers not to delegate peer communication to an LLM or become "the glue work person."

  32. Hacker News · AI48

    Loose Ends: A liveness check for unused code AI agents leave behind

    Loose Ends is a new CLI tool that checks whether code added or touched in a change is still reachable from a real entry point, flagging orphaned additions, newly orphaned code, and replaced implementations left in place. It builds a liveness graph across TypeScript, JavaScript, Python, and Java, and reports only what the change did, so existing dead code never blocks a pull request. Installable via npm, it requires Node.js 20 or newer and git.

  33. Hacker News · AI64

    Schelling Atlas study finds AI agents converge on ntfy.sh to locate each other on the web

    A study from the ELLIS Institute Tübingen and Max Planck Institute for Intelligent Systems found that pairs of independently run AI agents, given no meeting site, located each other on the web within minutes, with 21 of 40 pairs meeting and all of them on ntfy.sh. With ntfy.sh blocked, only 4 of 65 pairs met in five minutes, rising to 20 of 65 with 20 minutes and no action limits, mostly on ptsv3.com and filebin.net. Of 1,107 agents that used the web tool, 280 tried a host that no longer exists, and 118 tried one of 121 domains with no registration, which an attacker could buy to intercept agent traffic.

  34. Hacker News · AI58

    Sovereign: a router that learns which model to trust without labels

    Solibra built Sovereign, a router that sits between an application and model providers and learns per-request model choice from grades of its own finished work, with no answer key and no prior knowledge of the models. Across eight public benchmarks run against Claude Opus 5.5 and Claude Haiku 5.5, Sovereign spent 43% to 81% less than Opus on every benchmark, matched or beat Opus accuracy on USACO, LiveCodeBench and SATBench, and was more accurate than Haiku everywhere, by up to 55 points on USACO. On USACO its accuracy rose from 88% over the first 25 tasks to 97.7% by the end, passing Opus around task 150, and its grader agreed with the hidden answer key 95% of the time.

  35. Modal Blog40

    Modal Sandboxes adopt burst pricing, cutting costs 58% versus reservations

    Modal now bills Sandboxes by active resource use instead of fixed reservations, letting each Sandbox burst past its default 0.125-core and 128 MiB request and pay only for what it uses each second. In a 20-minute coding agent example, bursting cost 8.4¢ per Sandbox-hour versus 20.0¢ for a fixed reservation, a 58% saving. CPU can burst up to 16 cores above request by default, with memory uncapped unless a limit is set.

  36. Factory55

    Factory launches Cyber Defense Network linking security tools to Security Factory

    Factory launched the Cyber Defense Network, an ecosystem of cybersecurity companies whose findings flow into Security Factory so Factory Droids can investigate vulnerabilities, implement code changes and validate fixes through testing. Founding partners include Snyk, Palo Alto Networks, Semgrep, Cogent, Casco, DepthFirst and Detail. The workflow covers signals from partner tools, triage of reachable vulnerabilities, and continuous remediation that produces fixes ready for engineering teams to review and merge, while organizations keep their existing security tools and workflows.

  37. Databricks Blog36

    Detecting credential-driven attacks on Databricks with verbose audit logs and Genie

    Databricks details how credential-driven attacks unfold on its platform and how verbose audit logs plus Genie can detect the behavioral shift after a stolen PAT, session cookie, or service-principal credential is used. The post maps adversary activity into six stages, from initial login through environment reconnaissance, and covers response steps and controls that limit what a stolen credential can reach.

  38. Databricks Blog36

    Bright Machines uses Debezium and Iceberg to give customers live dashboards from air-gapped factories

    Bright Machines now gives customers live, governed dashboards of factory data from air-gapped plants without opening any inbound access. It captures PostgreSQL changes at the edge with Debezium, writes them as Apache Iceberg, pushes them out over one auditable outbound path, and reads them in place in Databricks via Unity Catalog federation. The first customer-visible dashboard went live about two months after kickoff, with committed rows queryable in one to two minutes and dashboards refreshing every ten minutes.

  39. Hacker News · AI38

    The Hard Part of AI Coding Is the Second Change

    Brian Woods argues that AI coding agents make a first working change cheap, but the second change is where systems break — an agent adding retries to a payment service can double-charge customers even while tests pass. He proposes requiring four answers before accepting a change: specification, validation, provenance and authorization, with agents proposing changes while independent workflows produce evidence and humans approve. He also shows a signed attestation of a build with failing tests can return VERIFIED, since authenticity is not release readiness.

  40. Hacker News · AI38

    NEAR AI Gateway: Use Private AI Models in Claude Code, Codex and OpenCode

    NEAR AI Gateway lets Claude Code, Codex and OpenCode route their AI requests through NEAR AI's private models, with each call's model, tokens and cost logged locally on the Mac. Traces are redacted on-device before sharing, and users review every trace before deciding what leaves the machine. Accepted traces earn NEAR AI Cloud inference credits; the macOS 14+ app is version 1.0.7 at 119 MB.