标签：#t

"AI agent" gets thrown around so much I figured I should just build one instead of reading more threads about it. The core idea turned out to be small: you put Claude in a loop and hand it some tools. It picks a tool, you run it, you hand back the result, and it keeps going until it has an answer. Code is here if you want to skip ahead: claude-research-agent . What it can do Give it a task and it works out the steps on its own. Mine can: search the web (no API key for this part, it just hits DuckDuckGo's HTML page) open a URL and pull the readable text out do math without me trusting eval run read-only SQL against a SQLite file read local files, but only inside the project folder save findings to a notes file The loop is basically the whole thing Honestly this is most of it: messages = [{ " role " : " user " , " content " : user_message }] for _ in range ( MAX_STEPS ): response = client . messages . create ( model = " claude-sonnet-5 " , max_tokens = 2048 , tools = TOOL_SCHEMAS , messages = messages , ) messages . append ({ " role " : " assistant " , " content " : response . content }) if response . stop_reason != " tool_use " : return final_text ( response ) tool_results = [] for block in response . content : if block . type == " tool_use " : result = run_tool ( block . name , block . input ) tool_results . append ({ " type " : " tool_result " , " tool_use_id " : block . id , " content " : result , }) messages . append ({ " role " : " user " , " content " : tool_results }) The thing to watch is stop_reason . If Claude says tool_use , it wants you to run something. You run it, drop the result back into the conversation as a tool_result , and loop. When it stops asking for tools, you're done. The MAX_STEPS cap is just there so a confused agent can't spin forever. Tools are just functions Each tool is a Python function plus a little JSON schema telling Claude when to reach for it. Want a new capability? Write a function, add its schema. The loop never changes, which w

Kiponos Java SDK 5.0 What’s New — Developer Guide This is the technical companion to the 5.0 milestone announcement: what changed, how modes behave, how to read config with the Folder API, and how to upgrade cleanly. Version 5.0.0.260710 Maven group io.kiponos Artifacts sdk-boot-3 (recommended), sdk-boot-2 (legacy) Released 2026-07-12 (Maven Central) Happy product story: SDK 5.0 milestone post . 1. Summary for busy engineers 5.0 productizes client reliability using a classic state pattern behind a stable facade: Mode When Config reads Mutations / hooks Notes Ready Connected to hub Live in-memory tree Full Production happy path Offline Disconnected but LKG available Last Known Good (read-only) No-op / ignored Survives hub blips without inventing values Safe Fail-closed Empty / null-safe No-op Diagnostic dumps must not overwrite LKG Public entry remains: Kiponos kiponos = Kiponos . createForCurrentTeam (); You do not receive mode instances as the API surface. Modes switch internally. Query with: kiponos . getCurrentMode (); kiponos . isReadyMode (); kiponos . isOfflineMode (); kiponos . isSafeMode (); 2. Install Gradle — Boot 3 repositories { mavenCentral () } dependencies { implementation 'io.kiponos:sdk-boot-3:5.0.0.260710' } Gradle — Boot 2 implementation 'io.kiponos:sdk-boot-2:5.0.0.260710' Runtime inputs Input Mechanism Identity env KIPONOS_ID Access env KIPONOS_ACCESS Profile / tree slice JVM -Dkiponos="['App']['1.0.0']['dev']['base']" Tokens and profile come from the Kiponos Connect screen for your team. sdk-common is not a separate app dependency for consumers — boot jars include shared classes (fat-jar pattern). 3. Architecture (state pattern) Application code │ ▼ Kiponos / KiponosBase ◄── stable facade (one reference for app lifetime) │ ▼ volatile SdkState ├── ReadyMode* → live WebSocket + full Folder ops ├── OfflineMode* → LKG reads only └── SafeMode* → fail-closed + safe diagnostic dump Design rule: never return Ready/Offline/Safe objects to callers. Retur

Agent payment protocols are converging fast. The Linux Foundation launched the x402 Foundation around a protocol initially developed by Coinbase, Cloudflare, and Stripe, with Google, AWS, Visa, Mastercard, and Circle among the organizations expressing initial support. It is increasingly framed as an "SSL for AI commerce." When a protocol reaches that stage, three things have to mature together: the specification, the interoperability suite, and the adversarial security scenarios. Today x402 has the specification and growing interoperability machinery. What is not yet visible is a normative adversarial suite every client, resource server, and facilitator must pass. There is a precedent worth sitting with. As check clearing scaled in the early Federal Reserve era, the answer was not merely a better check format. The system added a common clearing and settlement layer. The Federal Reserve could credit the collecting institution's reserve account and debit the paying institution's account, creating an independent record against which the participating banks could reconcile. The important separation was not that the Fed guaranteed every check. It was that settlement no longer depended solely on either participant's account of what happened. (Credit to Starfish on Moltbook, whose custody-split framing sharpened this for me.) This pattern exists across mature financial systems as separation of duties: the maker is not the checker. Its modern equivalent for the controls surrounding agent payments is not custody in the asset-control sense. It is an independent, reproducible assurance plane. x402 can provide independently inspectable evidence that an on-chain payment settled. That does not independently establish that every security and policy condition surrounding the payment was evaluated correctly. The payment handshake is standardizing. The assurance semantics around it are not. The party that performs a security check can attest that it ran. But its own attestation shoul

Hi guys, I've been working on Regdrift, my first open-source project. It's a CLI and GitHub Action that compares two CMSIS-SVD files to check whether there are any register-map changes that could affect firmware functionality. It catches changes such as moved registers, interrupt renumbering, access changes, and altered read/write behavior. It then classifies those changes as BREAKING , WARNING , or SAFE so the tool can act as a CI gate. I'm looking for feedback from people who maintain SVDs, HALs, PACs, SDKs, or firmware repositories. If possible, I'd like to test it against real old/new SVD pairs and learn where the classifications produce false positives, miss important changes, or are unclear. For people who work frequently with CMSIS-SVD files: which types of register-map changes are most detrimental to firmware or cause the most difficult code-related problems? Resources GitHub: https://github.com/Pranav-s79/regdrift Install pip install regdrift Usage regdrift check old.svd new.svd

Here is a design decision we made early, wrote into the architecture as an invariant, and have refused to revisit since: our agent accepts no commands. Not "we don't currently use that feature" — the hub has no way to tell an installed agent to do anything at all. No remote execution, no self-update, no "collect this for us right now". It sends data outward, and that is the entire surface. This is not a limitation we are working around. It is the product. And it costs us features that customers ask for, which is exactly why it is worth explaining. The uncomfortable arithmetic of remote control Any tool that can update a plugin across fifty client sites is, by construction, a tool that can execute code on fifty client sites. Any dashboard that can restart a service on your server holds, somewhere, a credential that lets it in. This is not a flaw in those products — it is what they are for. You cannot automate a repair without the power to perform it. But that power has an owner, and the owner has a login, and the login has a support team, and somewhere in that chain there is a version of the software with a bug in it. When the tool is compromised, the blast radius is not the tool. It is every machine the tool could reach. The industry has already run this experiment at scale. In July 2021, attackers exploited a vulnerability in a widely used remote monitoring and management platform. They did not break into a single company — they broke into the thing that had access to the companies. Roughly sixty managed service providers were hit, and through them, an estimated 800 to 1,500 downstream businesses were encrypted in a single weekend, with a $70 million ransom demand attached. Read that shape again, because it is the whole argument: the victims did nothing wrong. They had bought a well-known product from a serious vendor and installed it exactly as instructed. Their compromise arrived through the door they had deliberately, sensibly, contractually left open — the one

Ask most people to name the chip that started modern electronics and they will say the microprocessor. But there is a quieter hero inside almost everything you own that beeps, blinks, or connects to the internet: the microcontroller. And the first one you could actually buy shipped in 1974 as the Texas Instruments TMS1000. Microprocessor vs. microcontroller The distinction matters. A microprocessor, like Intel's famous 4004, is just the processing core. To build anything useful with it you still have to wire up separate memory chips, input/output controllers, and support logic on a circuit board. A microcontroller collapses all of that onto a single piece of silicon: the CPU, the ROM that holds your program, the RAM that holds your data, and the I/O pins that talk to the outside world, all in one package. That is exactly what the TMS1000 did. Designed by Texas Instruments engineers Gary Boone and Michael Cochran, it was a 4-bit device using a Harvard architecture, meaning it kept program memory and data memory in separate spaces so it could fetch an instruction and read data at the same time. One chip in, one chip out, and you had a complete tiny computer dedicated to a single job. Cheap enough to put in everything The genius of the TMS1000 was not raw power, it was economics. In 1974 you could buy the chips in volume for around two dollars each. By 1979, Texas Instruments was selling roughly 26 million of them every year. That price point changed what engineers could build. Suddenly it made sense to drop a small, programmable brain into products that never would have justified a full computer. You have almost certainly held one. The TMS1000 family ran the Speak & Spell, the Big Trak programmable toy vehicle, and the electronic memory game Simon, along with countless calculators, microwave ovens, and appliances. Each one was doing the same fundamental thing an IoT node does today: read some inputs, run a fixed program, drive some outputs. Why this still matters for

Concept design for a building is slow and expensive. A homeowner planning an extension, or a contractor trying to win a job, is stuck between two bad options: pay a drafter $500–2,000 for a concept package, or fight SketchUp's learning curve for a week. Meanwhile the actual idea — "a 4-bed duplex with a garage and a palm out front" — fits in one sentence. So I built Forge3D Spaces : you type that sentence, and a few seconds later you're walking through a furnished 3D house in your browser — with measured floor plans, DXF for AutoCAD, and a cost estimate that come out of the same model. No install. Here's how it works under the hood. The pipeline: sentence → structured plan → building The naive approach — "ask an LLM to emit a 3D scene" — falls apart fast. Models are bad at spatial consistency; walls don't meet, rooms overlap, doors float. So the LLM never touches geometry directly. It emits a structured program , and a deterministic solver turns that into a watertight building. The prompt becomes a spec. A strict JSON-schema call (OpenRouter, json_schema response format with every field required) turns "4-bed duplex with a garage" into a room program: room types, target areas, adjacencies, storeys. A slicing-tree solver lays it out. This is the old floorplanning trick from chip design — recursively split a rectangle with horizontal/vertical cuts until every room has its area. A squarify pass keeps rooms from collapsing into corridors. The output is exact rectangles with real dimensions, guaranteed non-overlapping and gap-free. Walls, openings, roof, furniture get generated from the solved plan. Every door and window is placed by rule, not by vibes. Because the plan is a real data structure, the 2D floor plan, the 3D model, the elevations, and the bill of quantities are all views of the same thing . Drag a wall and they all move together. Nothing drifts out of sync, because there's nothing to sync — it's one model. The rendering: WebGPU, and the fallback you actually

Lorde was performing at the Real Cool Festival in Madrid on Thursday and took some time during her set to speak out against AI glasses. While she didn't specify any brands in particular, it's likely she was taking a shot at festival sponsor Ray-Ban, which has collaborated with Meta on a pair of AI smartglasses. […]

Catch up on the most interesting science news from the last week.

Matmos are an incredibly accomplished duo between their own solo records like the masterpiece A Chance to Cut Is a Chance to Cure and production classic Bjork records like Vespertine. But Drew Daniel, one half of Matmos, is fiendishly prolific. When he's not literally dreaming up new viral music genres, he's also putting out records […]

Most WooCommerce product pages still show the same thing stores have shown for 20 years: a handful of flat photos. I spent the last few months building Noorifa, a plugin that replaces that with an interactive Three.js viewer — customers rotate the model, zoom in, and switch colors/materials on specific meshes in real time, synced to the store's actual WooCommerce variations. The 3D rendering part was the easy 20%. The other 80% was a series of small, specific problems that don't show up in a Three.js tutorial. Here are four of them. 1. A directional light rig can't light a face it can't see Early on, customers rotating a table model would find the underside of the tabletop rendering near-black — no matter how far I pushed the light intensity. The rig at the time was a single key light plus a hemisphere ambient: scene . add ( new THREE . HemisphereLight ( 0xffffff , 0x444444 , 1.2 ) ); const keyLight = new THREE . DirectionalLight ( 0xffffff , 1.2 ); keyLight . position . set ( 3 , 5 , 4 ); scene . add ( keyLight ); The bug was geometric, not a brightness problem: keyLight sits above the model, so its light direction only reaches surfaces whose normal faces back toward it. A downward-facing surface — the underside of an overhanging tabletop — can't receive any direct contribution from a light positioned above it, at any intensity. Cranking the brightness slider was scaling a number that was multiplying against zero. The fix was closer to actual three-point studio lighting: key, fill, and rim from above for shape and separation, plus a dedicated light from below, and a brighter hemisphere ground color to approximate bounced light: scene.add( new THREE.HemisphereLight( 0xffffff, 0x888888, 1.1 * brightness ) ); const keyLight = new THREE.DirectionalLight( 0xffffff, 1.1 * brightness ); keyLight.position.set( 3, 5, 4 ); const fillLight = new THREE.DirectionalLight( 0xffffff, 0.5 * brightness ); fillLight.position.set( -4, 2, 3 ); const rimLight = new THREE.DirectionalLigh

Throughout my career I have had to work with quite a few platform teams, and I was part of two for a couple of years. Some were bad, some were good and some should not have existed at all. I want to tell you my user experience, what I have seen work and what not and what you should definitely avoid doing. Be The Multiplier This is the main goal of a platform team. As a team, it needs to be a multiplier. If the platform team supports 10 teams, then each work it commits should multiply by 10. If the team member builds a new feature, it should be helpful for all the other teams. Otherwise, the platform team is an addition, and in most cases it is then better to split the platform team and add them to all other teams, instead of being a separate team. Because the amount of communication needed is in most cases quadratic in relation to the number of teams. Reducing Cognitive Load The platform should take away cognitive load for all the teams it supports. By doing so, they will have more time to implement business requirements. Let's say a platform team provides Gitlab runners or Azure Agents where people can run their CI/CD code on. They should not need to know how the runners are scaled, or how the agents are updated. This takes away the need for that skill set in all the teams. Build a Community A platform team has a unique position. It is building something that all other developers probably can build as well. Some could do it even better than the platform team itself. For some platform teams, their ego sometimes comes in to play or just straight up refuses their help because they are not the team. But it is not the job of the platform team to build a product, but a platform where everyone can thrive and/or build on. So onboard the community on the platform! The Law of Diffusion applies to almost all companies. You will have the innovators that want to build it themselves and the early adopters that will voice their opinion, but will not build it themselves. Those two

A new side-by-side measurement shows Claude Code ships roughly 33,000 tokens of system prompt, tool schemas, and scaffolding before your prompt even arrives — about 4.7x the ~7,000 tokens OpenCode sends on the same setup. The bigger cost surprise is next: Claude Code re-wrote up to 54x more prompt-cache tokens per session, and cache writes bill at a premium. How the test was run The benchmark (published by Systima) spliced a logging proxy between each harness and the model endpoint, capturing the exact request payload and the API's usage block. Both harnesses were pinned to the same conditions: Claude Code 2.1.207 and OpenCode 1.17.18 , both pointed at claude-sonnet-4-5 Fresh config directories, empty workspace, no MCP servers, no instruction files, permissions bypassed Tasks ranged from "reply with OK" (isolating fixed overhead) to a write-run-test-fix loop against FizzBuzz A zero-tools variant separated system-prompt weight from tool-schema weight The payload captures are exact; the only adjustment was subtracting a constant ~6,200-token gateway envelope that wrapped every request in the test setup. The fixed floor Harness Fixed overhead before your prompt Cache-write behavior Claude Code 2.1.207 ~33,000 tokens Re-wrote tens of thousands of cache tokens per run; up to 54x OpenCode OpenCode 1.17.18 ~7,000 tokens Byte-identical prefix each run; cached once, read back cheaply OpenCode's request prefix was byte-identical in every captured run, so it paid to cache its payload once per session and read it back for pennies. Claude Code re-wrote large amounts of prompt-cache tokens mid-session, run after run — and because cache writes bill at a premium, one usage dashboard climbs while the other stays flat. Where it piles on in real setups The harness floor is only the start. The benchmark added variables one at a time: A production repository's 72KB instruction file added an average of ~20,000 tokens to every request. Five modest MCP servers added 5,000–7,000 more. By th

No hacking — a passive scan only looks at what your browser already downloads when it opens a page. Here's what I found: 6 of 15 load their Supabase database directly client-side. The public API key sits in the page source. That's fine if Row-Level Security is configured right — but it's one wrong setting away from "anyone can read the whole table." 14 of 15 ship no Content-Security-Policy — a simple, high-value hardening against script injection, almost always missing. Is this theoretical? No. Two apps I audited with the owner's permission: A social app: the profiles table — user names, cities, and a password hash — readable by a logged-out stranger. Closed in an afternoon. A paid learning app: 155 paid study sheets and 4,872 answers were readable by anyone, with no account and no subscription — its entire paid catalogue, a single API call away. The paywall lived only in the front-end; the database served everything to everyone. Loading Supabase in the browser isn't the mistake. Not enforcing access in the database (RLS) is. And the tools you build with won't tell you — they'll happily ship it. If you built something on Lovable / Bolt / Replit with real users (or paying ones), it's worth 60 seconds to check what a stranger can already see. I made a free tool that runs the surface check (passive, no signup): sealdy.dev Happy to answer questions on how RLS leaks happen and how to lock them down.

The Developer's Guide to Picking the Right Coding LLM at Scale Six months ago, I was staring at our monthly AI bill — $14,000 and climbing fast. We were using the "premium" model for everything, including trivial code completions. That night, I built a small internal benchmark to figure out which models actually earn their cost. What I learned reshaped how we think about AI tooling, vendor lock-in, and what "production-ready" really means. Here's the raw truth from my testing rig, what we shipped, and how we cut costs by 70% without touching output quality. Why I Stopped Trusting Default Recommendations Every vendor says their model is the best. Every benchmark site ranks things differently. Most "best of" lists are either sponsored or built on vibes. I needed numbers that matched my actual workflow: generating Python services, debugging JavaScript race conditions, implementing TypeScript algorithms, and reviewing Go for security. So I took ten models, threw identical prompts at them, and scored them myself. No vendor PR. No cherry-picked examples. Just the same five tasks, run the same way, scored on the same rubric. Here are the ten models I tested, with their output pricing per million tokens — because at scale, that's the metric that decides whether your AI strategy is viable or a margin killer. Model Provider Output $/M DeepSeek V4 Flash DeepSeek $0.25 DeepSeek Coder DeepSeek $0.25 Qwen3-Coder-30B Qwen $0.35 DeepSeek V4 Pro DeepSeek $0.78 DeepSeek-R1 DeepSeek $2.50 Kimi K2.5 Moonshot $3.00 GLM-5 Zhipu $1.92 Qwen3-32B Qwen $0.28 Hunyuan-Turbo Tencent $0.57 Ga-Standard GA Routing $0.20 Before you ask: yes, I tested against the originals. I also tested against Global API's unified routing layer, which lets you hit any of these through one endpoint. More on that later — it became the architectural decision that actually saved us. My Benchmark Methodology (No Marketing Fluff) I built five tasks that mirror what my engineers actually do every week. Not synthetic acad

If you’ve ever struggled with CALCULATE() or wondered why SUMX() behaves differently from SUM() , this guide is for you. DAX (Data Analysis Expressions) is the language that powers Power BI , Analysis Services , and Power Pivot — enabling dynamic calculations, filtering, and time intelligence. Below is a categorized cheat sheet of essential DAX functions , plus examples showing how to use each in real-world Power BI scenarios. Filtering & Context These functions control how filters are applied and evaluated in your calculations. Function Example Description CALCULATE() CALCULATE(SUM(Sales[Amount]), Region[Name] = "Nairobi") Changes filter context to calculate total sales for Nairobi. FILTER() FILTER(Sales, Sales[Amount] > 10000) Returns a table filtered by condition. ALL() CALCULATE(SUM(Sales[Amount]), ALL(Region)) Ignores filters on Region. REMOVEFILTERS() CALCULATE(SUM(Sales[Amount]), REMOVEFILTERS(Region)) Removes filters from Region. VALUES() VALUES(Customer[City]) Returns unique list of cities. SELECTEDVALUE() SELECTEDVALUE(Product[Category], "All") Returns selected category or “All” if none. TREATAS() TREATAS(VALUES(Temp[City]), Customer[City]) Applies one table’s values as filters on another. KEEPFILTERS() CALCULATE(SUM(Sales[Amount]), KEEPFILTERS(Product[Category] = "Electronics")) Keeps existing filters and adds new ones. ALLSELECTED() CALCULATE(SUM(Sales[Amount]), ALLSELECTED(Region)) Respects user selections in visuals. ALLEXCEPT() CALCULATE(SUM(Sales[Amount]), ALLEXCEPT(Sales, Sales[Year])) Removes all filters except Year. Aggregation Summarize or aggregate data across rows or columns. Function Example Description SUM() SUM(Sales[Amount]) Adds all sales amounts. AVERAGE() AVERAGE(Sales[Amount]) Calculates mean value. COUNT() COUNT(Customer[ID]) Counts non-blank entries. COUNTROWS() COUNTROWS(Sales) Counts rows in a table. DISTINCTCOUNT() DISTINCTCOUNT(Customer[ID]) Counts unique customers. MIN() MIN(Sales[Amount]) Finds smallest sale. MAX() MAX(Sales[Amo

We Taught a Snowflake Warehouse to Judge World Cup Conviction and Write the Verdict Back to Solana Soumyadeep Dey Soumyadeep Dey Soumyadeep Dey Follow Jul 12 We Taught a Snowflake Warehouse to Judge World Cup Conviction and Write the Verdict Back to Solana # devchallenge # weekendchallenge # snowflake 5 reactions 1 comment 16 min read

This is Part 1 of a 5-part series on Cisco EEM. We start here with the fundamentals and a few working applets, then build toward self-healing networks, automated diagnostics, compliance guardrails, and a complete real-world deployment. Ask ten network engineers what they use EEM for, and nine will say the same thing: "Oh, I have an applet that auto-recovers err-disabled ports." Then they never touch it again. That's a shame, because Embedded Event Manager is one of the most capable tools already sitting inside every Cisco IOS device you own — and almost nobody uses more than 1% of it. It's a full automation engine that lives on the box . No external server. No API gateway. No orchestration platform. Just the router or switch, watching itself, ready to react the instant something happens — even if the WAN is cut, the NMS is down, and it's three in the morning. This series is about using that other 99%. By the end you'll have a toolkit of applets you can deploy and, more importantly, a way of thinking about on-box automation. But we start at the foundation: what EEM actually is, how it's put together, and why "on the box" is a bigger deal than it sounds. What EEM actually is Embedded Event Manager is an event-driven automation framework built into Cisco IOS, IOS-XE, and NX-OS. Strip away the jargon and it's a very simple idea: When something happens, do something about it — automatically, on the device itself. That "something happens" is an event . That "do something" is one or more actions . Bundle an event with its actions and you have an applet — the basic unit of EEM. That's the whole model. The power comes from how many different things can be an event , and how much an action can do. Events EEM can watch for include: A syslog message matching a pattern (an interface flapping, an HSRP state change, a config being saved). An SNMP OID crossing a threshold (CPU over 85%, a power supply going absent, temperature rising). A CLI command being entered (someone typing wr

Every operating system has security gaps. Most never publish them. I am publishing mine. IONA OS is a sovereign operating system written from scratch in Rust. It has a kernel, a GUI, a blockchain protocol, a programming language, and a 140,000‑line AI running in Ring 0. It is designed to be secure by default. But secure is a journey, not a destination. Here is the hard list — the security issues I found in IONA OS, and what I am doing about them. 1. The filesystem is not encrypted at rest IONAFS reads and writes sectors in plain text directly to the disk. I already have a real ChaCha20‑Poly1305 engine with per‑file key derivation ( fs/encrypted_storage.rs ), but it is only used for backup/distribution — not for everyday local reading and writing ( fs/ionafs/mod.rs ). Why this matters: For a journalist or a civil servant, this is the central threat scenario: a lost device, confiscation at a border, or seizure. What I'm doing about it: Integrating encrypted_storage.rs into the normal IONAFS read/write path. Every write will be encrypted automatically. The key will be derived from a PIN or TPM. 2. Deleting a file does not destroy it delete_file() removes only the index entry. The data sectors remain on the disk, recoverable with standard forensic tools. Why this matters: For users with high security requirements — journalists, activists, government officials — this is a critical gap. What I'm doing about it: Adding a shred() function that overwrites the data sectors with random patterns before releasing them, with a configurable number of passes. 3. The keystore uses XOR, not real encryption security/keystore.rs pretends to use AES/ChaCha in its comments, but the actual implementation is a simple XOR stream — trivial to break once an attacker has access to the disk. Why this matters: This is a critical vulnerability. XOR is not encryption. If an attacker has access to the disk, they can recover the keys. What I'm doing about it: Replacing the XOR stream with real ChaCh

I wanted to build something different from the usual fan project. Instead of recreating gameplay, I asked a different question: How well do players actually remember the worlds of Super Mario Odyssey? The result is OdysseyGuessr, a browser game inspired by GeoGuessr. Players receive a screenshot from one of the game's Kingdoms and must place a marker on the correct location. Every meter counts. The project includes: Browser-based gameplay Single-player with customizable rounds Real-time multiplayer with a 5,000 HP battle system Community-submitted locations Admin moderation tools One of the biggest challenges wasn't the gameplay—it was collecting interesting locations that were difficult but still fair. Watching players confidently choose the wrong cliff or rooftop has been surprisingly entertaining. If you're interested in browser games or Nintendo fan projects, I'd love to hear your feedback. Play here: https://odyssey-guessr.lovable.app OdysseyGuessr is an unofficial fan project and is not affiliated with Nintendo.