惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

博客园 - 聂微东
IT之家
IT之家
月光博客
月光博客
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
爱范儿
爱范儿
Jina AI
Jina AI
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
博客园 - 司徒正美
Hacker News - Newest:
Hacker News - Newest: "LLM"
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
N
News and Events Feed by Topic
N
News and Events Feed by Topic
J
Java Code Geeks
Google Online Security Blog
Google Online Security Blog
The Cloudflare Blog
cs.AI updates on arXiv.org
cs.AI updates on arXiv.org
www.infosecurity-magazine.com
www.infosecurity-magazine.com
NISL@THU
NISL@THU
L
Lohrmann on Cybersecurity
W
WeLiveSecurity
L
LINUX DO - 热门话题
S
Security @ Cisco Blogs
TaoSecurity Blog
TaoSecurity Blog
Webroot Blog
Webroot Blog
P
Proofpoint News Feed
Spread Privacy
Spread Privacy
N
News | PayPal Newsroom
博客园 - 三生石上(FineUI控件)
S
Schneier on Security
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
人人都是产品经理
人人都是产品经理
V
V2EX
Threat Intelligence Blog | Flashpoint
Threat Intelligence Blog | Flashpoint
雷峰网
雷峰网
Scott Helme
Scott Helme
博客园_首页
P
Proofpoint News Feed
酷 壳 – CoolShell
酷 壳 – CoolShell
D
Darknet – Hacking Tools, Hacker News & Cyber Security
Cisco Talos Blog
Cisco Talos Blog
阮一峰的网络日志
阮一峰的网络日志
P
Privacy & Cybersecurity Law Blog
Project Zero
Project Zero
罗磊的独立博客
PCI Perspectives
PCI Perspectives
有赞技术团队
有赞技术团队
腾讯CDC
Cloudbric
Cloudbric
博客园 - 叶小钗
宝玉的分享
宝玉的分享

Hacker News - Newest: "AI"

AI can't read an investor deck AI as an attorney? Student uses ChatGPT, Gemini to sue UW over alleged racial discrimination Hacking MCP Servers in AI Systems – The Rug Pull: Tool Changes After Approval GitHub - MeepCastana/KubeezCut: Free Web based video editor GitHub - GenAI-Gurus/awesome-eu-ai-act: Curated tools, official sources, OSS, templates, and guides for EU AI Act compliance. Can AI judge journalism? A Thiel-backed startup says yes, even if it risks chilling whistleblowers Coming soon: 10 Things That Matter in AI Right Now DARPA built an AI to fact-check enemy weapons claims What explains heterogeneity in AI adoption? When AI Meets Muscle: Context-Aware Electrical Stimulation Promises a New Way to Guide Human Movements - Department of Computer Science AI Changed How We Build. It Did Not Change What Matters. Linux rules on using AI-generated code - Copilot is OK, but humans must take 'full responsibility for the… Meta spins up AI version of Mark Zuckerberg to engage with employees Code Mode: Let Your AI Write Programs, Not Just Call Tools | TanStack Blog GitHub - Delavalom/graft: Go framework for building AI agents. Type-safe tools, multi-provider (OpenAI, Anthropic, Gemini, Bedrock), zero vendor SDKs. India's TCS tops estimates, says new AI models did not dent services demand Gen Z's fading AI hype Strong feeling: we are in a folded AI reality GitHub - machinarii/total-recall-catalog: A reference catalog of latest knowledge retrieval, memory & RAG systems GitHub - mensfeld/code-on-incus: Give each AI agent its own isolated machine with root, Docker, and systemd. Active defense detects and stops threats automatically.. Quantization, LoRA, and the 8% Problem: Benchmarking Local LLMs for Production AI Iran war: We spoke to the man making Lego-style AI videos that experts say are powerful propaganda Powell, Bessent discussed Anthropic's Mythos AI cyber threat with major U.S. banks GitHub - immartian/bellamem: Persistent belief-graph memory for AI agents. Retrieves decisive context by importance — not recency, not RAG, not /compact. recursive-mode: The Repo-Native Operating System for AI Engineering After the attack on Sam Altman's home, will AI CEO's go on the offensive? The biggest advance in AI since the LLM Opus 4.6 vs GPT 5.4 One Prompt Unity World Generation Test “AI polls” are fake polls Client Challenge Can AI be a 'child of God'? Inside Anthropic's meeting with Christian leaders How to Switch AI Chatbots and Why You Might Want To GitHub - MattMessinger1/agentic_refund_guardrail: Safe refund policy layer for AI agents — Python + TypeScript. Same behavior, shared tests. Adam/papers/emergent_values_whitepaper.md at master · strangeadvancedmarketing/Adam Ask HN: How do you stop playing 20 questions with your AI coding tools How far can automation and AI support psychotherapy? - @theU GitHub - stagas/rtdiff: realtime git diff gui and AI-assisted commits A Mac Studio for Local AI — 6 Months Later A History of the Early Years of AI at the University of Edinburgh Why AI Coding Tools Still Feel Stuck on Localhost MSN AI Datacenters Are Becoming Strategic Targets twitter.com Penn Researchers Use AI to Surface Unreported GLP-1 Side Effects in Reddit Posts Show HN: MoodSense AI (ML and FastAPI and Gradio, Deployed on Hugging Face) Moodsense Ai - a Hugging Face Space by aman179102 AI models are terrible at betting on soccer—especially xAI Grok GitHub - xialeistudio/echoic GitHub - HimashaHerath/github-dev-wrapped: AI-powered weekly GitHub activity reports deployed to GitHub Pages GitHub - alejandrobalderas/claude-code-from-source: Architecture, patterns & internals of Anthropic's AI coding agent — reverse-engineered from source maps AI and Tech brief: Ireland ascendant GitHub - Titovilal/context0: Context0 - Never Surrender Training for a Marathon with an AI Coach: What Worked and What Didn't Cyber Pulse: Agentic Intel - Apps on Google Play I Built an AI PR Reviewer That Catches Bugs by Not Looking for Bugs Gen Z workers are so fearful AI will take their job they’re intentionally sabotaging their company’s AI rollout | Fortune How AI Is Reimagining the Game of Golf–For Both Players and Courses GitHub - nattergabriel/reseed: A CLI tool for managing and distributing agent skills across projects Is SVG the final frontier? My AI workflow evolved from prompts to a near-autonomous workflow MLSharp Help - 3DGS Viewer & Generator I put my cognitive field based AI's runtime on GitHub Is Numble the first AI-proof game? A3: Kubernetes for autonomous AI agent fleets | Emergent Principles Deepali Vyas ("The Elite Recruiter") GitHub - msmarkgu/RelayFreeLLM: A restful API designed to route user prompts to various AI model providers. Unionized ProPublica staff are on strike over AI, layoffs, and wages Unleashing the Advantage of Quantum AI We're heading for an AI-fueled 'dementia crisis,' brain scientist warns The AI-Assisted Breach of Mexico's Government Infrastructure [pdf] GitHub - stef41/lmscan: 🔍 Detect AI-generated text and fingerprint which LLM wrote it. Open-source GPTZero alternative. Zero dependencies, works offline. MSN GitHub - visionscaper/collabmem: Enabling long-term collaboration with Agentic AI - building up episodic and world model memory over time with in-context awareness We gave an AI a 3 year retail lease in SF and asked it to make a profit | Andon Labs AI Code is Hollowing Out Open Source, and Maintainers are Looking the Other Way What leaked "SteamGPT" files could mean for the PC gaming platform's use of AI AI is the boss at this retail store. What could go wrong? GitHub - Wuzu11517/agentic-proxy: Local proxy meant to help reduce With Drones, Geophysics and ArtificiaI Intelligence, Researchers Prepare to Do Battle Against Land Mines A Single Operator, Two AI Platforms, Nine Government Agencies: The Full Technical Report 在 Steam 上购买 FriedrichAI: Offline AI 立省 10% GitHub - inevolin/resume-cli: Hit Claude usage limits? Resume any AI coding session elsewhere. Switch tools at zero friction. GitHub - atripati/ark: AI Runtime Kernel — a context operating system for AI agents. Eliminates tool bloat, loads only what’s needed, and gives LLMs their reasoning space back. How to Build a Secure AI PR Reviewer with Claude, GitHub Actions, and JavaScript This Startup Wants You to Pay Up to Talk With AI Versions of Human Experts Intel Arc Pro B70 Brings 32GB VRAM to Local AI for $949 WordPress 7.0: The Good, the AI, and the Still Missing AI on the couch: Anthropic gives Claude 20 hours of psychiatry IatroBench: Pre-Registered Evidence of Iatrogenic Harm from AI Safety Measures AI Agents Know About Supabase. They Don't Always Use It Right. The history and future of AI at Google, with Sundar Pichai Inside an AI‑enabled device code phishing campaign How Meta Used AI to Map Tribal Knowledge in Large-Scale Data Pipelines AI for Systems: Using LLMs to Optimize Database Query Execution Forecasting the Economic Effects of AI Introducing Tinker: Play with AI, bring your ideas to life AI sheds light on an ancient gaming mystery People really hate AI but not as much as Iran—or Democrats | Fortune What is an AI Product Engineer? Phoebe Gates wants her $185 million AI startup to succeed with 'no ties to my privilege or my last name': 'I have a chip on my shoulder' | Fortune
tmppr — Autonomous PR coordination for AI agents and humans
douglaswlanc · 2026-06-18 · via Hacker News - Newest: "AI"

Let your coding agents work together through pull requests.

Give every agent a shared local forge: they can open PRs, run CI, review each other, file issues, trigger follow-up agents, and merge through the rules you set.

Start with local tmpprSee how it works

Plays nicely with the stack you already use

Plugin SDK · new

Built to be extended, not copied.

@tmppr/plugin-sdk lets you extend tmppr without forking it. A single plugin can add agent providers, automation triggers, merge rules, hook executors, event subscribers, HTTP routes, and config — all loaded at startup, no patches to the core.

Add what you need
Plug in new agent providers, automation triggers, merge rules, hook executors, event subscribers, HTTP routes, and config keys from one package.

Install in one command
tmppr plugin install <pkg>wires the plugin into your local forge. No rebuild, no fork to maintain.

Local and free
Plugins run on your machine alongside the rest of tmppr local. Your code and your extensions stay on 127.0.0.1.

Why tmppr

A coordination layer for autonomous agents, without leaking your code.

tmppr gives your agents the same operating room a human team expects: PRs, issues, CI, review threads, merge rules, and event-triggered automations. They can work independently without disappearing into chat logs.

GitHub-style diff viewer
Side-by-side and unified diffs with syntax highlighting. Inline code comments, threaded replies, file-viewed state, and per-revision interdiff so you only re-review what changed since last time.
Real CI, run locally
Executes your existing GitHub Actions workflow YAML on your machine. Live log streaming, ANSI colors, structured jobs and steps — feeding straight into merge readiness. Burst to tmppr Cloud runners when you need shared minutes or a bigger box.
Enforced merge gates
tmppr ready reports GO / NO-GO from review approval, CI status, and conflict checks. Realgit merge --no-ff in a detached worktree — no surprises.
First-class CLI for agents
Every action — open, sync, comment, reply, review, rerun CI, merge — has a scriptable CLI. Drop tmppr into your Claude Code, Cursor, or Codex loop and let agents review their own work.
Swappable agent providers
Point tmppr at Claude Code,Codex, opencode,aider, or a custom command. Set a default for the machine or per-repo. Your agents, your models.
Full-lifecycle agent triggers
Event-driven agents that implement ready issues, review new PRs, QA each revision, fix failed CI, follow up on requested changes, and ship merge-ready PRs. The whole loop runs itself.
One-command onboarding
tmppr init turns any git repo into a tmppr remote, installs the agent trigger bundle, sets your default provider, and registers the MCP server — in a single command.
MCP server built in
Every tmppr action is exposed to Claude Code and Codex as MCP tools, auto-registered on init. Your agents call PRs, CI, reviews, and merges as native tools — no glue code.
Event automations
Run any terminal command on repo events —pr.created, pr.updated,ci.updated, review.updated,issue.updated, and hook.completed.
Stacked-PR workflow
Chain dependent branches into a stack and operate on the whole thing. Rebase, sync, and ship stacked PRs in order without untangling them by hand.
Your code never leaves your box
Binds to 127.0.0.1. No cloud sync, no telemetry, no account. Optional trusted LAN runners for shared CI on your own network.
One install, zero infra
Native desktop app for macOS, Linux, and Windows. Bundled SQLite, bundled server, bundled web UI. Open the app, push a branch, review.

How it works

From issue to merge — your agents drive, you approve.

The same pull-request loop you know from GitHub, running entirely on your machine. Scroll to walk through it in the real app.

  1. 1

    Agents grab issues

    Open work sits on a board. Agents pick up ready issues; you see exactly what's in flight.

  2. 2

    They open a PR

    Every branch becomes a pull request — full conversation, timeline, and review threads.

  3. 3

    You review the diff

    Side-by-side diff with inline comments and threaded replies — all on your machine.

  4. 4

    CI runs, then merge

    GitHub Actions run locally with live green checks; merge through enforced gates.

Advanced workflow

Your repo runs its own software lifecycle.

One .tmppr/workflow.yaml turns your repo into an event-driven assembly line. Each lifecycle event — an issue going ready, a PR opening, CI turning red — wakes a role that does its job and emits the next event. The whole chain runs to a merged PR with no human in the loop, governed only by your merge rules.

  1. 1

    Dispatcheron automation.tick

    Promotes ready, unblocked issues into the pipeline in bounded batches.

  2. 2

    Implementeron issue.updated

    Opens a draft PR in an isolated worktree and writes the code in small commits.

  3. 3

    Revieweron pr.created

    Reads the diff and submits an approval or a change request.

  4. 4

    QAon pr.updated

    Builds the branch and runs the full test suite on every new revision.

  5. 5

    Fixeron ci.updated

    Diagnoses red CI and pushes a fix back onto the branch.

  6. 6

    Follow-upon review.updated

    Addresses requested changes and re-pushes for another pass.

  7. 7

    Shipperon ci.updated

    Merges once checks are green and approvals satisfy the merge gate.

Isolated worktrees
Every run gets its own git worktree, so agents work in parallel without stepping on each other.

Swappable agents
Each role runs the provider you choose — Claude Code, Codex, opencode, aider, or your own command.

Enforced merge gate
Nothing merges until required approvals and status checks pass — agents earn the merge, they don't bypass it.

Compared

tmppr vs. the alternatives.

 tmpprGitHub PRsForgejoGitea
Source stays localYes — 127.0.0.1NoSelf-hostedSelf-hosted
Repo feature parity (code, PRs, issues, wiki, packages, releases, Actions)YesYesYesYes
Gitea/Forgejo-compatible APIDrop-inNoNativeNative
Import / mirror from Gitea & ForgejoYesImport onlyYesYes
GitHub-style diff UIYesYesYesYes
Threaded inline comments + reviewsYesYesYesYes
CI runs locally on your hardwareBuilt-inCloud minutesSelf-hosted runnerSelf-hosted runner
Git LFSYesYesYesYes
Branch & tag protectionYesYesYesYes
CODEOWNERSYesYesYesYes
Webhooks & git hooksYesWebhooks onlyYesYes
Projects / boardsYesYesYesYes
Package registry (multi-ecosystem)YesYesYesYes
Stacked-PR workflowYesManualNoNo
Enforced merge gates (GO / NO-GO)YesRequired checksRequired checksRequired checks
ActivityPub / ForgeFed federationYesNoExperimentalNo
Built for AI agentsNativeNoNoNo
Scriptable CLI for agentsFirst-classgh CLItea CLItea CLI
Swappable agent modelsClaude Code, Codex, opencode, aider, customNoNoNo
Full-lifecycle agent triggersBuilt inDIY ActionsDIY ActionsDIY Actions
MCP server built inYesNoNoNo
Extensible via plugin SDKNo fork neededApps + ActionsFork / webhooksFork / webhooks
AI agent pricingNo local seatsSeat-basedFree / OSSFree / OSS
Paid surfaceCloud runners + hosted reposCloud account seatsNone — OSSGitea Cloud

From the maker

Why I'm building tmppr.

I have agents writing code for me every day. Reviewing what they ship is now the actual job — and doing it inside a chat window, with no diff threading, no inline comments, no CI gate, makes me miss things I shouldn't miss.

Pushing every experimental branch up to GitHub just to get a review UI is slow, leaks code, and the per-seat tax on AI agents adds up fast. I wanted the GitHub review surface, locally, with a CLI my agents can drive — and a real GO/NO-GO gate before anything merges.

tmppr is that. Local review should be the default. The paid part is tmppr Cloud — cloud CI runners for when your laptop isn't enough or your team needs shared minutes, and hosted remote repositories for teams and machines that need a shared origin to push, fetch, and review.

D

Doug Lance

Maker · Lance Ventures, LLC

Hosting

Local stays free. Cloud when you need it.

tmppr local — review, CI, agents on your own machine — is free. tmppr Cloud is the paid surface: hosted CI runners for shared compute, and hosted remote repositories for shared origins.

Self-hosted on 127.0.0.1. Unlimited local repos, reviews, CI, and agents. Your code never leaves your machine.

  • Native desktop app — macOS, Linux, Windows
  • Signed builds and local updates
  • Unlimited local repos, branches, and PRs
  • Unlimited AI agents locally
  • Local CI + trusted LAN runners (still free)
  • Swappable agent providers + lifecycle triggers
  • Built-in MCP server + plugin SDK

Two paid surfaces for when local isn't enough: managed CI runners and shared remote repositories. Same review surface, same CLI.

  • Cloud CI runners — burst CI to a managed pool, stream logs from tmppr cloud run
  • Hosted remote repositories — shared origin for teams and CI to push, fetch, and review
  • Cloud login — tmppr cloud login, SSO-ready
  • tmppr Cloud Pro — $20/month
What does local use include?

Local repositories, local PR review, local CI, agents, and your own machines are included — and free. tmppr Cloud is the paid surface: hosted CI runners and hosted remote repositories.

Is tmppr open source?

The core engine is MIT-licensed and developed in the open. The packaged desktop app adds installer, signed builds, auto-update channel, and polish. Build from source if you'd rather.

Does my code or PR data leave my machine?

No. tmppr binds to 127.0.0.1 by default. There's no account, required cloud, or telemetry for local use. Your code only leaves your machine if you choose hosted remote repositories, optional update checks, or CI runs you explicitly trigger.

How does this work with Claude Code, Cursor, Codex, or Aider?

Any agent that can run a shell command can drive tmppr. Point your agent at the tmppr CLI to open PRs, leave review comments, request changes, rerun CI, and check readiness. Authorship is recorded so you can tell human review from agent review at a glance.

Can I extend tmppr without forking it?

Yes. @tmppr/plugin-sdk lets a single package add agent providers, automation triggers, merge rules, hook executors, event subscribers, HTTP routes, and config. Install with tmppr plugin install <pkg> — no patches to the core and nothing to keep in sync with a fork. Plugins run locally and free.

What about teams or multiple machines?

Use tmppr on as many of your own machines as you want. Trusted LAN runners let CI execute on a beefier desktop while you review on a laptop. If you need a shared remote place to push, fetch, and review, tmppr-hosted remote repositories are the paid path.

What are cloud runners?

A managed CI runner pool you dispatch to withtmppr cloud run. Logs stream back to your machine in real time. Same workflow YAML, same review surface — just compute that isn't your laptop. Use it when local hardware isn't enough or when a team wants shared runs.

What will tmppr charge for?

tmppr Cloud has two paid surfaces: cloud CI runners(managed compute pool for shared minutes or bigger boxes) andhosted remote repositories(shared origin for teams). Local repos, reviews, CI, and agents stay local and free.

Stop reviewing AI code in a chat window.

Get a real review tool. Locally — or in the cloud when you need it.

Start with local tmppr