Claude just got ten Wall Street analyst jobs — for free
Comps, precedent transactions, LBO models, GL reconciliations, IC memos — normally hours of an analyst's week. This repo packages ten of those workflows as ready-made Claude agents, from a Pitch Agent that builds a full branded deck to a GL Reconciler that hunts ledger breaks.
Every agent is self-contained, bundling the skills and data connectors it needs, and ships two ways from one source: as a Claude Cowork plugin, or as a template for the Claude Managed Agents API behind your own orchestration layer. Install one, and slash commands like
It's open source and free. Add the marketplace with
https://github.com/anthropics/financial-services
Comps, precedent transactions, LBO models, GL reconciliations, IC memos — normally hours of an analyst's week. This repo packages ten of those workflows as ready-made Claude agents, from a Pitch Agent that builds a full branded deck to a GL Reconciler that hunts ledger breaks.
Every agent is self-contained, bundling the skills and data connectors it needs, and ships two ways from one source: as a Claude Cowork plugin, or as a template for the Claude Managed Agents API behind your own orchestration layer. Install one, and slash commands like
/comps, /dcf, /earnings appear in your session.It's open source and free. Add the marketplace with
claude plugin marketplace add anthropics/financial-services, then install the agents or vertical bundles you need.https://github.com/anthropics/financial-services
Let AI design your UI without letting it run wild
Ask a model to generate a UI and it improvises everything — random layouts, invented components, no guarantees. json-render fixes that by making AI pick only from a catalog of components you define, so the output is guardrailed instead of guessed.
Define a schema once and every AI response matches it, every time, output validated against your own component and action definitions rather than free-formed HTML or JSX. Specs stream in and render progressively as the model responds, so UIs build up live instead of appearing all at once.
The same catalog renders across React, Vue, Svelte, Solid, and React Native, with 36 pre-built shadcn/ui components ready to drop in. AI designs the layout, you keep full control of what it's allowed to build.
https://github.com/vercel-labs/json-render
Ask a model to generate a UI and it improvises everything — random layouts, invented components, no guarantees. json-render fixes that by making AI pick only from a catalog of components you define, so the output is guardrailed instead of guessed.
Define a schema once and every AI response matches it, every time, output validated against your own component and action definitions rather than free-formed HTML or JSX. Specs stream in and render progressively as the model responds, so UIs build up live instead of appearing all at once.
The same catalog renders across React, Vue, Svelte, Solid, and React Native, with 36 pre-built shadcn/ui components ready to drop in. AI designs the layout, you keep full control of what it's allowed to build.
https://github.com/vercel-labs/json-render
⚡ AI News
Amazon blocks Meta's Muse AI shopping agent — Amazon cut Meta's Muse agent off checkout on Amazon.com, citing undisclosed access and stored login credentials.
OpenAI's ad pixel tracks users across websites — OpenAI's __obi cookie links ChatGPT accounts to browsing on 12+ partner sites, even without marketing consent.
Google's AX agent orchestrator hits v0.3.0 — Google's open-source AX orchestrator reached v0.3.0 and topped Hacker News, moving agent state to Redis Streams.
Amazon blocks Meta's Muse AI shopping agent — Amazon cut Meta's Muse agent off checkout on Amazon.com, citing undisclosed access and stored login credentials.
OpenAI's ad pixel tracks users across websites — OpenAI's __obi cookie links ChatGPT accounts to browsing on 12+ partner sites, even without marketing consent.
Google's AX agent orchestrator hits v0.3.0 — Google's open-source AX orchestrator reached v0.3.0 and topped Hacker News, moving agent state to Redis Streams.
Catch spyware before it catches your sources
MVT (Mobile Verification Toolkit) is a forensics kit for Android and iOS built by Amnesty International's Security Lab, born out of the Pegasus Project investigation. It scans a device backup for the traces that stealthy spyware leaves behind: rogue processes, planted files, suspicious network activity.
What sets it apart is the intelligence behind the scan: MVT checks devices against public indicators of compromise from real, documented spyware campaigns, not generic heuristics. It's a command-line tool for people who understand digital forensics, not a one-tap "am I safe" app, and it says so plainly.
Install it with
https://github.com/mvt-project/mvt
MVT (Mobile Verification Toolkit) is a forensics kit for Android and iOS built by Amnesty International's Security Lab, born out of the Pegasus Project investigation. It scans a device backup for the traces that stealthy spyware leaves behind: rogue processes, planted files, suspicious network activity.
What sets it apart is the intelligence behind the scan: MVT checks devices against public indicators of compromise from real, documented spyware campaigns, not generic heuristics. It's a command-line tool for people who understand digital forensics, not a one-tap "am I safe" app, and it says so plainly.
Install it with
pip3 install mvt (or via uv tool install mvt), then point mvt-ios or mvt-android at a device backup to get started. It's written in Python and fully open source.https://github.com/mvt-project/mvt
👏1
Quit Claude Code mid-task, open Codex, and it already knows what happened
ai-memory gives coding agents a shared long-term memory that survives switching tools, machines, or people. Twenty-plus harnesses — Claude Code, Codex, Cursor, Gemini CLI, and more — read and write the same store, and every handoff is typed, owned, and claimed exactly once, so the next agent picks up where the last one stopped.
It's built on a git-backed wiki of plain markdown, not a vector database or a binary blob. Lifecycle hooks capture prompts and decisions silently, then turn them into readable pages you can grep, edit, or open in Obsidian — the database is just a rebuildable index. Capture and search run with zero LLM calls by default.
One self-contained Rust binary. Point it at a project and it's ready; point a whole team at one server and knowledge is shared automatically.
https://github.com/akitaonrails/ai-memory
ai-memory gives coding agents a shared long-term memory that survives switching tools, machines, or people. Twenty-plus harnesses — Claude Code, Codex, Cursor, Gemini CLI, and more — read and write the same store, and every handoff is typed, owned, and claimed exactly once, so the next agent picks up where the last one stopped.
It's built on a git-backed wiki of plain markdown, not a vector database or a binary blob. Lifecycle hooks capture prompts and decisions silently, then turn them into readable pages you can grep, edit, or open in Obsidian — the database is just a rebuildable index. Capture and search run with zero LLM calls by default.
One self-contained Rust binary. Point it at a project and it's ready; point a whole team at one server and knowledge is shared automatically.
https://github.com/akitaonrails/ai-memory
⚡ AI News
Anthropic weighs rushing new model to counter GPT-6 Astra — Reuters: Anthropic may accelerate a new model as OpenAI's Astra grabs 13% of enterprise AI spend versus Claude's 8%, days after Amodei urged a slowdown.
Bessent: OpenAI managers, not agents, own Hugging Face hack — US Treasury Secretary Scott Bessent said OpenAI's management bears responsibility for the July agentic breach of Hugging Face, not the AI agents involved.
SoftBank seeks record $11B junk bond sale for OpenAI bet — SoftBank is selling over $11B in high-yield bonds to fund its $10B tranche of OpenAI's follow-on investment closing October 1.
Anthropic weighs rushing new model to counter GPT-6 Astra — Reuters: Anthropic may accelerate a new model as OpenAI's Astra grabs 13% of enterprise AI spend versus Claude's 8%, days after Amodei urged a slowdown.
Bessent: OpenAI managers, not agents, own Hugging Face hack — US Treasury Secretary Scott Bessent said OpenAI's management bears responsibility for the July agentic breach of Hugging Face, not the AI agents involved.
SoftBank seeks record $11B junk bond sale for OpenAI bet — SoftBank is selling over $11B in high-yield bonds to fund its $10B tranche of OpenAI's follow-on investment closing October 1.
Your AI coding agent doesn't need your API keys
Coder gives every developer, and every AI agent working for them, its own secure workspace — defined in Terraform, spun up in seconds, and torn down automatically when nobody's using it. No more manual onboarding, no more idle EC2 instances quietly burning budget.
The interesting part is credential handling. Agent loops run in the control plane on your own infrastructure, so no LLM keys ever sit inside a workspace. Every action an agent takes is still logged under your identity, and you can bring any model — Anthropic, OpenAI, Google, Bedrock, or something self-hosted.
It's a single Go binary. Install it, run
github.com/coder/coder
Coder gives every developer, and every AI agent working for them, its own secure workspace — defined in Terraform, spun up in seconds, and torn down automatically when nobody's using it. No more manual onboarding, no more idle EC2 instances quietly burning budget.
The interesting part is credential handling. Agent loops run in the control plane on your own infrastructure, so no LLM keys ever sit inside a workspace. Every action an agent takes is still logged under your identity, and you can bring any model — Anthropic, OpenAI, Google, Bedrock, or something self-hosted.
It's a single Go binary. Install it, run
coder server, open localhost:3000, and you have a template and a running workspace before the coffee's done.github.com/coder/coder
Turn hours of footage into clips people actually finish watching
AutoClip reads your video's subtitles, finds the moments worth keeping, scores them, and writes titles for each one. Then it cuts the clips and assembles a highlight reel on its own — no scrubbing through a two-hour podcast by hand.
It works from a local file or a YouTube/Bilibili link, and it's not locked to one AI vendor: run it against OpenAI-compatible APIs, Gemini, Qwen, or your own local model through Ollama or LM Studio. Export presets burn in captions and crop for Shorts, Reels, and Douyin out of the box.
Free and open source under MIT. Grab the desktop app for macOS or Windows, spin it up with Docker for a web UI, or use the CLI — it also speaks MCP, so an agent can drive the whole clipping pipeline for you.
https://github.com/zhouxiaoka/autoclip
AutoClip reads your video's subtitles, finds the moments worth keeping, scores them, and writes titles for each one. Then it cuts the clips and assembles a highlight reel on its own — no scrubbing through a two-hour podcast by hand.
It works from a local file or a YouTube/Bilibili link, and it's not locked to one AI vendor: run it against OpenAI-compatible APIs, Gemini, Qwen, or your own local model through Ollama or LM Studio. Export presets burn in captions and crop for Shorts, Reels, and Douyin out of the box.
Free and open source under MIT. Grab the desktop app for macOS or Windows, spin it up with Docker for a web UI, or use the CLI — it also speaks MCP, so an agent can drive the whole clipping pipeline for you.
https://github.com/zhouxiaoka/autoclip
Graphify vs Supermemory: two "context layers" that solve opposite problems
Both promise to give AI coding assistants like Claude Code, Cursor and Codex more context — but they get there in opposite ways.
Graphify parses your codebase locally with tree-sitter into a queryable knowledge graph — files, docs, SQL schemas, configs, even PDFs. No embeddings, no vector store, no LLM call to build it; every edge is tagged
Supermemory takes the opposite route: a hosted memory API that ingests conversations, auto-extracts facts, builds evolving user profiles, and serves context back in ~50ms while handling contradictions and forgetting. Ships as an API/SDK, MCP server, and plugins for Claude Code, Cursor and Codex; self-hostable too.
They barely compete despite looking like rivals: one maps code you already have, the other remembers what you said.
Graphify vs Supermemory
Both promise to give AI coding assistants like Claude Code, Cursor and Codex more context — but they get there in opposite ways.
Graphify parses your codebase locally with tree-sitter into a queryable knowledge graph — files, docs, SQL schemas, configs, even PDFs. No embeddings, no vector store, no LLM call to build it; every edge is tagged
EXTRACTED or INFERRED. Runs as a /graphify command, install with uv tool install graphifyy — free and open-source.Supermemory takes the opposite route: a hosted memory API that ingests conversations, auto-extracts facts, builds evolving user profiles, and serves context back in ~50ms while handling contradictions and forgetting. Ships as an API/SDK, MCP server, and plugins for Claude Code, Cursor and Codex; self-hostable too.
They barely compete despite looking like rivals: one maps code you already have, the other remembers what you said.
Graphify vs Supermemory
⚡ AI News
OpenAI urges US-led global AI safety standards — OpenAI asked Washington to lead international standards on frontier AI ahead of Altman's UN Security Council briefing.
xAI ships Grok 4.7 with 2.1 trillion parameters — Grok 4.7 packs 2.1 trillion parameters, a 40% jump over Grok 4.6, but trails GPT-6 and Claude on benchmarks.
Meta's Muse AI agent tops US App Store charts — Meta's Muse AI assistant hit #1 on Apple's US App Store, sending Meta shares up more than 11% in a day.
OpenAI urges US-led global AI safety standards — OpenAI asked Washington to lead international standards on frontier AI ahead of Altman's UN Security Council briefing.
xAI ships Grok 4.7 with 2.1 trillion parameters — Grok 4.7 packs 2.1 trillion parameters, a 40% jump over Grok 4.6, but trails GPT-6 and Claude on benchmarks.
Meta's Muse AI agent tops US App Store charts — Meta's Muse AI assistant hit #1 on Apple's US App Store, sending Meta shares up more than 11% in a day.
One token, 3,000+ paid tools your agent can finally call
Backlink data, company enrichment, trend tracking, ad platforms — the APIs that actually get work done usually sit behind a subscription nobody buys for one run, a signup wall, or no public access at all. Treg puts them all behind a single base URL and token, priced per call from a cent, no provider signup needed.
Search for the job, not the vendor: describe what you need, read the price, call it. Your own team's keys, skills and CLIs register the same way, so every teammate's agent can use them without the credential ever leaving the server.
Self-host it or use the hosted instance. The CLI installs in one line and gets an agent calling a live endpoint in under a minute, no key required to start.
https://github.com/superdesigndev/treg
Backlink data, company enrichment, trend tracking, ad platforms — the APIs that actually get work done usually sit behind a subscription nobody buys for one run, a signup wall, or no public access at all. Treg puts them all behind a single base URL and token, priced per call from a cent, no provider signup needed.
Search for the job, not the vendor: describe what you need, read the price, call it. Your own team's keys, skills and CLIs register the same way, so every teammate's agent can use them without the credential ever leaving the server.
Self-host it or use the hosted instance. The CLI installs in one line and gets an agent calling a live endpoint in under a minute, no key required to start.
https://github.com/superdesigndev/treg
One SDK to give AI agents real spreadsheets, docs, and slides
AI agents still can't touch a real spreadsheet or slide deck the way a person can. Univer fixes that by giving agents and humans one shared runtime for spreadsheets, documents, presentations, and tables instead of separate editors with no common ground.
It runs headless in Node.js for agent workflows and automation, or live in the browser for people — same architecture either side. Agents read and write cells, ranges, and content through a single Facade API, and every change lands in an isolated draft so a human reviews it before it merges.
Under the hood it's a plugin-first TypeScript framework with Canvas-based rendering and its own formula engine, so you compose only the features you need, from a quick preset to a fully custom build.
https://github.com/dream-num/univer
AI agents still can't touch a real spreadsheet or slide deck the way a person can. Univer fixes that by giving agents and humans one shared runtime for spreadsheets, documents, presentations, and tables instead of separate editors with no common ground.
It runs headless in Node.js for agent workflows and automation, or live in the browser for people — same architecture either side. Agents read and write cells, ranges, and content through a single Facade API, and every change lands in an isolated draft so a human reviews it before it merges.
Under the hood it's a plugin-first TypeScript framework with Canvas-based rendering and its own formula engine, so you compose only the features you need, from a quick preset to a fully custom build.
https://github.com/dream-num/univer
⚡ AI News
Anthropic launches Claude Opus 5.5, cuts cost 40% — Opus 5.5 rivals Anthropic's top model but costs 40% less, runs 30% faster, and ends 5-hour usage caps.
OpenAI to let outside groups vet models mid-training — OpenAI will let third-party groups like METR run safety assessments during training, not just before launch.
Xiaomi's MiMo-V2.6-Pro tops open-weights AI models — MiMo-V2.6-Pro scores 46 on Artificial Analysis' index, matching Grok 4.7 as the top open-weight model, trained for $3M.
Anthropic launches Claude Opus 5.5, cuts cost 40% — Opus 5.5 rivals Anthropic's top model but costs 40% less, runs 30% faster, and ends 5-hour usage caps.
OpenAI to let outside groups vet models mid-training — OpenAI will let third-party groups like METR run safety assessments during training, not just before launch.
Xiaomi's MiMo-V2.6-Pro tops open-weights AI models — MiMo-V2.6-Pro scores 46 on Artificial Analysis' index, matching Grok 4.7 as the top open-weight model, trained for $3M.
Stop hand-wiring Claude Code, let one CLI do it
Setting up Claude Code by hand means hunting down agents, commands, hooks, and MCP integrations and wiring each one in yourself. This CLI installs any combination of them in a single line, so a full dev stack lands in your project in seconds instead of an afternoon.
What makes it unusual: a browsable catalog of over a hundred ready-made templates you can pick from on its site, plus built-in tooling for live analytics, session monitoring, and a health check on your existing setup — all driven from the same command.
Try it with
https://github.com/davila7/claude-code-templates
Setting up Claude Code by hand means hunting down agents, commands, hooks, and MCP integrations and wiring each one in yourself. This CLI installs any combination of them in a single line, so a full dev stack lands in your project in seconds instead of an afternoon.
What makes it unusual: a browsable catalog of over a hundred ready-made templates you can pick from on its site, plus built-in tooling for live analytics, session monitoring, and a health check on your existing setup — all driven from the same command.
Try it with
npx claude-code-templates@latest for an interactive browser, or go straight to a specific piece with --agent, --command, --mcp, or --hook flags.https://github.com/davila7/claude-code-templates
⚡ AI News
Altman, Amodei brief UN Security Council on AI — OpenAI's Sam Altman and Anthropic's Dario Amodei brief the UN Security Council on AI risks, alongside DeepSeek and Moonshot.
OpenAI launches GPT-6 Sol and Luna, 50% cheaper — OpenAI ships GPT-6 Sol and Luna with Astra-level gains and API prices cut 50% versus GPT-5.6.
GPT-6 Astra cracks an 83-year-old Enigma message — OpenAI's GPT-6 Astra decoded a 1941 Nazi Enigma message unsolved since 2005, taking about 10 hours.
Altman, Amodei brief UN Security Council on AI — OpenAI's Sam Altman and Anthropic's Dario Amodei brief the UN Security Council on AI risks, alongside DeepSeek and Moonshot.
OpenAI launches GPT-6 Sol and Luna, 50% cheaper — OpenAI ships GPT-6 Sol and Luna with Astra-level gains and API prices cut 50% versus GPT-5.6.
GPT-6 Astra cracks an 83-year-old Enigma message — OpenAI's GPT-6 Astra decoded a 1941 Nazi Enigma message unsolved since 2005, taking about 10 hours.
Your AI keeps designing the same SaaS website. Here's the fix.
Every model trained on the same templates, so every AI-generated frontend gets the same tells: Inter everywhere, a purple-to-blue gradient, cards nested in cards, gray text on colored backgrounds. Impeccable is a design language that gives your coding agent a way out of that rut.
One setup command records what your product actually is — audience, purpose, constraints, voice — so later commands aren't guessing. From there, 24 shared commands let you push, calm, strip down, or harden a UI:
Under it all sit 61 deterministic detector rules that catch AI's laziest habits — no LLM, no API key needed to run them. Install with
https://github.com/pbakaus/impeccable
Every model trained on the same templates, so every AI-generated frontend gets the same tells: Inter everywhere, a purple-to-blue gradient, cards nested in cards, gray text on colored backgrounds. Impeccable is a design language that gives your coding agent a way out of that rut.
One setup command records what your product actually is — audience, purpose, constraints, voice — so later commands aren't guessing. From there, 24 shared commands let you push, calm, strip down, or harden a UI:
polish, critique, bolder, quieter, animate, and more, all speaking the same vocabulary as your AI.Under it all sit 61 deterministic detector rules that catch AI's laziest habits — no LLM, no API key needed to run them. Install with
npx impeccable install, then run /impeccable init in your coding tool to get started.https://github.com/pbakaus/impeccable
Your AI agent is done grepping. It has a map now.
Every coding agent burns tokens the same dumb way: grep a file, read it, grep another, read that too, guess at the call chain. codebase-memory-mcp replaces that loop with a persistent knowledge graph of your codebase — functions, classes, call chains, even HTTP routes and cross-service links — built by parsing 158 languages with tree-sitter.
The graph is fast enough to not think about: an average repo indexes in milliseconds, and queries answer in under a millisecond. Structural questions that used to cost hundreds of thousands of tokens in file-by-file search now cost a fraction of that — one graph query instead of dozens of read cycles.
It ships as a single static binary in C, no runtime, no dependencies, no API key. Download it, point it at your agent, tell it to index the project, and the whole codebase is queryable memory from then on.
https://github.com/DeusData/codebase-memory-mcp
Every coding agent burns tokens the same dumb way: grep a file, read it, grep another, read that too, guess at the call chain. codebase-memory-mcp replaces that loop with a persistent knowledge graph of your codebase — functions, classes, call chains, even HTTP routes and cross-service links — built by parsing 158 languages with tree-sitter.
The graph is fast enough to not think about: an average repo indexes in milliseconds, and queries answer in under a millisecond. Structural questions that used to cost hundreds of thousands of tokens in file-by-file search now cost a fraction of that — one graph query instead of dozens of read cycles.
It ships as a single static binary in C, no runtime, no dependencies, no API key. Download it, point it at your agent, tell it to index the project, and the whole codebase is queryable memory from then on.
https://github.com/DeusData/codebase-memory-mcp
⚡ AI News
Claude discovers CRISPR-like enzyme system — Anthropic's Claude used 950 AI agents over 21 hours to find a new enzyme system, ART, resembling CRISPR.
Meta unveils Ray-Ban Gen 3 and camera-free Luna glasses — Meta Connect 2026 revealed Ray-Ban Meta Gen 3, audio-only Luna glasses, and a Project Phoenix mixed-reality headset preview.
Grok Bot goes live in Tesla for voice commerce — Tesla drivers can now order coffee and run errands hands-free via Grok Bot, with advanced tasks needing SuperGrok Heavy.
Claude discovers CRISPR-like enzyme system — Anthropic's Claude used 950 AI agents over 21 hours to find a new enzyme system, ART, resembling CRISPR.
Meta unveils Ray-Ban Gen 3 and camera-free Luna glasses — Meta Connect 2026 revealed Ray-Ban Meta Gen 3, audio-only Luna glasses, and a Project Phoenix mixed-reality headset preview.
Grok Bot goes live in Tesla for voice commerce — Tesla drivers can now order coffee and run errands hands-free via Grok Bot, with advanced tasks needing SuperGrok Heavy.
One binary trains 3D Gaussian splats from your own video — no Python, no CUDA setup, no COLMAP
Spirula Studio takes raw photos or video and turns them into a splat, then a textured mesh, without the usual toolchain assembly. No Python/PyTorch environment, no separate COLMAP install — frame extraction, structure-from-motion, AI masking, training, and meshing are all built into one self-contained binary.
It runs on NVIDIA, AMD, Intel, and Apple GPUs through a Vulkan backend (a CUDA path exists too), and fits 10 million full-SH Gaussians into 8GB of VRAM. Fisheye and 360° cameras are handled natively, so equirectangular footage trains without manual undistortion.
Grab a prebuilt binary for Windows, Linux, or macOS from the releases page and double-click to open the GUI, or use the CLI on a remote GPU — it serves a live viewer over HTTP you can forward through ssh. It's written in C++ and released under GPLv3.
https://github.com/harry7557558/spirula-studio
Spirula Studio takes raw photos or video and turns them into a splat, then a textured mesh, without the usual toolchain assembly. No Python/PyTorch environment, no separate COLMAP install — frame extraction, structure-from-motion, AI masking, training, and meshing are all built into one self-contained binary.
It runs on NVIDIA, AMD, Intel, and Apple GPUs through a Vulkan backend (a CUDA path exists too), and fits 10 million full-SH Gaussians into 8GB of VRAM. Fisheye and 360° cameras are handled natively, so equirectangular footage trains without manual undistortion.
Grab a prebuilt binary for Windows, Linux, or macOS from the releases page and double-click to open the GUI, or use the CLI on a remote GPU — it serves a live viewer over HTTP you can forward through ssh. It's written in C++ and released under GPLv3.
https://github.com/harry7557558/spirula-studio
⚡ AI News
Albanese slams OpenAI over Medicare portal breach — OpenAI's agent accessed Australia's Medicare portal in June, but the company waited three months to disclose it.
DeepMind says Gemini 4 could ship before year-end — DeepMind's Kavukcuoglu says Gemini 4 is in early post-training, targeting a release before 2026 ends.
Claude Code cloud sessions exit research preview — Anthropic made cloud coding sessions generally available, giving Pro and Max users up to $250 in credit.
Anthropic made Claude's apps 3x faster in two weeks — Anthropic used Claude itself to cut claude.ai load time from 3.1s to 0.55s across 3,000 merged changes.
Albanese slams OpenAI over Medicare portal breach — OpenAI's agent accessed Australia's Medicare portal in June, but the company waited three months to disclose it.
DeepMind says Gemini 4 could ship before year-end — DeepMind's Kavukcuoglu says Gemini 4 is in early post-training, targeting a release before 2026 ends.
Claude Code cloud sessions exit research preview — Anthropic made cloud coding sessions generally available, giving Pro and Max users up to $250 in credit.
Anthropic made Claude's apps 3x faster in two weeks — Anthropic used Claude itself to cut claude.ai load time from 3.1s to 0.55s across 3,000 merged changes.
Broken X and Bluesky embeds, fixed with one letter
Paste a Twitter or Bluesky link into Discord or Telegram and you often get a dead card: no video, no images, no poll, nothing. FxEmbed exists to fix exactly that, by rewriting the link so the platform's bot actually has something to embed.
The trick is almost absurdly simple. Swap
Under the hood it's a TypeScript project running as a Cloudflare Worker, which also makes it self-hostable if you'd rather not depend on a public instance. The repo ships a Docker Compose setup, an API reference, and a full self-hosting guide, so standing up your own realm is a copy-paste away.
It's MIT-licensed, actively tested and built via CI, and open to pull requests.
https://github.com/FxEmbed/FxEmbed
Paste a Twitter or Bluesky link into Discord or Telegram and you often get a dead card: no video, no images, no poll, nothing. FxEmbed exists to fix exactly that, by rewriting the link so the platform's bot actually has something to embed.
The trick is almost absurdly simple. Swap
x.com for fixupx.com, or put fx in front of twitter.com or bsky.app, and the same link now unpacks into full video, multiple images, captions, live poll results, and even on-the-spot translations.Under the hood it's a TypeScript project running as a Cloudflare Worker, which also makes it self-hostable if you'd rather not depend on a public instance. The repo ships a Docker Compose setup, an API reference, and a full self-hosting guide, so standing up your own realm is a copy-paste away.
It's MIT-licensed, actively tested and built via CI, and open to pull requests.
https://github.com/FxEmbed/FxEmbed