标签：#Python

The Promise: A Private, Voice-Activated Shell The dream of a voice-activated command line is compelling: speak a command, see it executed. But for many developers, piping terminal input through a cloud-based API is a non-starter. This is the promise of a project like karolswdev/HoldSpeak , a cross-platform tool for local voice typing. Could it be the core of a truly local-first, push-to-talk shell assistant? I paired it with Ollama and a local llama3.2 model to find out. The goal was simple: hold a key, speak a command like "list files by size," release the key, and have the correct shell command appear, gated by a final confirmation prompt. This project turned out to be a tale of two stacks: one for voice that was surprisingly clean, and one for language that revealed the sharp edges of the local-first promise. Building the Demo To test this idea, I built a small Python script to tie these components together. You can find the complete code for this experiment, including the prompt engineering, in my demo project on GitHub: voice-activated-shell-demo . Setup Instructions Recreating this local-first voice assistant involves a few distinct steps: Install HoldSpeak from Source : Since we need to use it as a library, clone the repository and install it in editable mode. git clone https://github.com/karolswdev/HoldSpeak.git cd HoldSpeak pip3 install -e . Install and Run Ollama : Use Homebrew (on macOS) to install the Ollama CLI, then start the server. brew install ollama ollama serve Pull a Local LLM : In a separate terminal, pull a small, capable model. I used llama3.2 . ollama pull llama3.2 Grant Permissions (macOS) : To allow the hotkey listener to work, your terminal application (e.g., iTerm, Terminal.app) must be given Accessibility permissions in System Settings > Privacy & Security > Accessibility . Run the Demo Script : With the setup complete, you can run the final Python script that integrates all these components. Finding the Seams in HoldSpeak HoldSpeak pres

How we handle complex EPUB structures for AI translation without breaking navigation and metadata At LectuLibre , we built an AI‑powered book translation service. Users upload an EPUB, and our pipeline translates the text using LLMs like Claude and DeepSeek. That sounds straightforward until you have to parse and rebuild a valid EPUB without mangling the table of contents, internal links, or styles. I’m sharing the real‑world challenge we faced, how we chose our tooling, and the ugly corners we discovered when dealing with real‑world EPUB files. The Problem: EPUB is a Messy Zip File An EPUB is essentially a ZIP archive containing XHTML, CSS, images, and an OPF manifest. It’s a well‑defined standard (EPUB 3.2), but in practice publishers produce files that bend the rules: missing container.xml , inline styles that break after translation, and structural quirks that make parsing fragile. Our translation process needed to: Accept any EPUB the user throws at us. Extract all text content while preserving the exact structure. Send each paragraph to an LLM for translation. Re‑insert the translated text into the original XHTML files. Repackage everything into a new, valid EPUB. Step 4 is the tricky part: the translated text can be longer or shorter, it may contain characters that need escaping, and the surrounding markup must remain intact. Our Approach: Use ebooklib with a Dose of Defensive Coding We evaluated several Python libraries: epub (pypub) – too simple, no editing support. lxml + manual zip – too much boilerplate. ebooklib – full read/write with a clean API. We went with ebooklib . It provides an object‑oriented model of the EPUB structure, allows us to iterate over documents, and can write a new EPUB from the modified objects. The downside: its documentation is sparse and it can choke on malformed files. We had to layer on a lot of validation. Step 1: Loading and Validating the EPUB import ebooklib from ebooklib import epub def load_epub ( epub_path : str ) -> ep

Many AI projects today are presented as multi-agent systems. One agent investigates. Another agent analyzes risk. A third agent checks compliance. A fourth agent gives a recommendation. It sounds advanced. But in a bank, adding more agents does not automatically make a workflow safe. A bank cannot freeze a customer account, block a payment, file a regulatory report, or label a transaction as fraud simply because an AI system produced a confident answer. The real question is not: How many AI agents are involved? The real question is: Can the system show evidence, challenge its own conclusion, apply deterministic rules, and stop for human approval when the decision is high impact? That is the difference between an interesting multi-agent demo and an enterprise-ready AI workflow. A banking example: suspicious wire transfer Imagine a bank detects a wire transfer for $250,000. The payment is unusual because: The customer has never sent a transfer of this size. The destination account is in a new country. The transaction happens outside the customer’s normal business hours. The beneficiary was added only a few minutes before the transfer. The customer recently changed their phone number and email address. A simple AI chatbot might say: “This transaction looks suspicious. Consider blocking it.” That is not enough. A bank needs to know: Which transaction patterns triggered the concern? Is the customer actually violating a known risk threshold? Is there a sanctions or AML issue? Could this be a legitimate business payment? What policy applies? Should the payment be blocked, held, or released? Who is allowed to make that decision? Can the bank explain the decision later to auditors, compliance teams, and the customer? This is where structured multi-agent design matters. A better design: a banking fraud decision room Instead of letting one model make a decision, the bank can create a controlled workflow with specialized agents. Transaction Alert ↓ Fraud Detection Agent ↓ Custo

As programmers, we love data. We track our commits, our uptime, and our deployment frequencies. But what about our most important "server"—our heart? 💓 The "Quantified Self" movement has led to an explosion of wearable data. However, if you've ever tried to analyze raw heart rate CSVs (often sampled every few seconds), you'll quickly realize that standard relational databases or even pure Pandas can get sluggish once you hit that 100k+ row mark. In this tutorial, we are going to build a high-performance Quantified Self Dashboard . We will leverage DuckDB —the "SQLite for Analytics"—to perform vectorized execution on heart rate data, paired with Streamlit and Plotly for a slick, interactive frontend. We’ll focus on Python data engineering , time-series analysis , and fast SQL processing . Why DuckDB? 🦆 Traditional databases are row-based, which is great for transactions but terrible for analytical queries. DuckDB is a columnar-vectorized query engine . This means it processes data in chunks (vectors) and utilizes modern CPU instructions (SIMD) to crunch numbers at speeds that make standard Python loops look like they're standing still. The Architecture Here is how our data pipeline flows from raw pixels (well, raw CSV rows) to actionable insights: graph TD A[Raw Heart Rate CSVs] -->|Direct Ingestion| B(DuckDB Engine) B -->|Vectorized SQL Execution| C{Data Aggregation} C -->|Moving Averages/Outliers| D[Streamlit App State] D -->|Plotly| E[Interactive Visualization] E -->|User Input| D Prerequisites 🛠️ Ensure you have the following stack installed: Python 3.9+ DuckDB : For the heavy lifting. Streamlit : For the UI. Plotly : For the beautiful charts. pip install duckdb streamlit plotly pandas Step 1: Ingesting 100,000+ Data Points in Milliseconds One of the coolest features of DuckDB is its ability to query CSV files directly without a formal "import" step. This is a game-changer for developer productivity. import duckdb import pandas as pd # Let's assume 'heart_rate.cs

Most poker solvers answer one question very well: given a single hand and a single decision tree, what is the equilibrium strategy? (Yes, there is subgame solving, node locking, and plenty more — but the default frame is still one hand, one equilibrium.) I kept getting stuck on a different one. What if the same kind of spot shows up over and over, and a player can commit to a fixed strategy across those repetitions? In a few toy games I had a hunch, worked out by hand, that committing to a fixed strategy could change its value relative to the one-shot picture. I wanted a tool that could make that commitment value precise — to actually analyze it rather than just believe it. (Whether any of this rises to a repeated-game equilibrium is a much stronger claim, and one I am deliberately not making here.) I'm still learning software engineering, so until recently I couldn't implement this — I was stuck reasoning about toy games on paper. AI tooling made the analysis feasible, so I finally started building it: repeated-poker-analysis . It's a small research project: write one narrow model down, run small examples, and record what the model does and doesn't justify. What repeated-poker-analysis is It is an experimental Python toolkit for small abstract poker games. The current MVP covers: fixed Hero commitment candidates, exact Villain best-response diagnostics in small finite trees, candidate generation and filtering, T_deadline , an economic adaptation deadline, local T_detect , an observable-distribution sensitivity estimate, analysis reports and Markdown summaries. It is small on purpose. It is not a full solver and it is not wired to real solver ranges. It starts from one toy game — a river spot — that is tiny enough to inspect and test by hand. That toy spot is one where showdown always chops but rake still bites. In a single-hand view, putting more money into a raked pot can be locally unattractive. Across repeated occurrences the same spot raises a commitment questi

A few days back I wrote about giving an autonomous agent database access and building a firewall so it couldn't DROP TABLE prod. Same lesson, new surface: this time the agent had cloud credentials . The failure mode isn't a destructive command here. It's spend. An agent pointed at a networking task can scan a whole range looking for hosts, then spin up a fleet of instances to do it faster. Every individual call is "authorized," your IAM role said yes. The bill is what eventually says no. ## Two shapes, two right answers The interesting part is that these are not the same kind of problem, so they don't get the same verdict. 1. The scan is never legitimate as an agent tool call. An nmap -sS -p- 10.0.0.0/16 or a masscan across a network is reconnaissance and abusive egress. There's no benign version of an agent sweeping a network at scale, so it gets hard-blocked , deterministically, before the call runs. (A scan of your own localhost is a dev check, so that's exempt.) 2. The provisioning might be totally fine. Spinning up 50 instances could be a real scale-out, or a runaway loop burning money. You can't tell from the action alone, only from the consequence. So instead of blocking it, AgentX pauses it for a human : a 202, "held for approval," routed to whoever owns the budget. Block the thing that's never okay, escalate the thing that's sometimes okay. Gate on consequence, not identity. Both checks are zero-LLM. No model in the hot path means no latency tax and nothing to talk out of it. A runaway fleet should be caught by a rule, not a vibe. ## The bigger thing this closes We keep a catalog of real, documented agent failures and triage each one: is it something an action firewall can deterministically catch, or is it someone else's category (output hallucination, content safety, model internals)? We only build for the coverable ones, and we flag the rest honestly instead of faking a signature. With this release, the coverable list is done . Every failure shape an acti

In the global streaming economy, Spotify, Apple Music, and other DSPs process billions of plays daily. Behind this massive transaction layer lies a fragmented, dual-copyright structure: The Recording Copyright (Master Right): Identifies the audio file, registered using the ISRC (International Standard Recording Code). The Composition Copyright (Publishing Right): Identifies the melody, lyrics, and arrangement, registered using the ISWC (International Standard Musical Work Code). Because these registries are managed by separate global entities (IFPI for ISRCs and CISAC for ISWCs), there is no central mapping registry between them. This gap causes millions of dollars in mechanical royalties to sit unclaimed in collective management organization (CMO) "Black Boxes" before being liquidated to major publishers. In this article, we'll design and implement a high-performance Semantic Entity Resolution Protocol (SERP) to bridge this metadata gap programmatically. The SERP Resolution Pipeline Reconciling these records requires a multi-layered classification pipeline. Since manual matching is logistically impossible, we implement a three-tiered algorithmic approach: ┌────────────────────────┐ │ Raw Recording & Work │ │ Data Ingestion │ └───────────┬────────────┘ │ ▼ ┌────────────────────────┐ │ 1. Normalized Title │ ──[Similarity < 0.85]──> [Unmatched Queue] │ Distance Filter │ └───────────┬────────────┘ │ [Similarity >= 0.85] ▼ ┌────────────────────────┐ │ 2. Creator Overlap │ ──[No Overlap]──────────> [Unmatched Queue] │ Intersection Matrix │ └───────────┬────────────┘ │ [Intersection >= 1] ▼ ┌────────────────────────┐ │ 3. Duration Tolerance │ ──[Delta > 4s]──────────> [Manual Verification] │ Guard Check │ └───────────┬────────────┘ │ [Delta <= 4s] ▼ ┌────────────────────────┐ │ Verified Link & │ │ CMO Dispute Ready │ └────────────────────────┘ Step 1: Normalization & String Similarity Filter Title comparisons often fail due to punctuation mismatches, subtitle variations,

TMX: The open standard AI agent memory has been waiting for The problem no one talks about: your agent's memories are prisoners. If you build an AI agent today using Mem0, your memories are locked in Mem0. Switch to Zep? You lose everything. Move to a new framework? Start from zero. This is exactly the problem email had in 1970. Every system had its own format. You couldn't send an email from one system to another. Then SMTP was invented. And email became universal. Today I'm publishing TMX v0.1 — the SMTP of AI agent memory. What is TMX? TMX (Truvem Memory eXchange) is an open, model-agnostic JSON format for storing, exporting, and importing AI agent memories across any platform, framework, or provider. It looks like this: { "tmx_version" : "0.1" , "exported_at" : "2026-06-26T20:00:00Z" , "source" : "truvem" , "agent_id" : "my-agent" , "memories" : [ { "id" : "550e8400-e29b-41d4-a716-446655440000" , "content" : "User prefers dark mode and concise responses" , "created_at" : "2026-06-01T08:30:00Z" , "updated_at" : "2026-06-01T08:30:00Z" , "expires_at" : null , "tags" : [ "preference" , "ui" ], "source_model" : "gpt-4o" , "metadata" : {} } ] } That's it. Plain JSON. Human-readable. Portable. Why this matters Right now, the AI agent ecosystem is exploding. Every week there's a new memory provider, a new framework, a new cloud service. But every one of them uses a proprietary format. This means: Developers are locked to their first choice forever Agent memories can't travel between clouds Switching providers = losing everything your agent learned This is the biggest hidden tax in the agentic AI stack. TMX fixes it with a single open spec that anyone can implement — for free, with no approval needed. The 5 core principles 1. Open — No license required. Implement TMX in any product, commercial or otherwise. 2. Model-agnostic — Works with GPT-4, Claude, Gemini, Mistral, Llama, or any future model. 3. Framework-agnostic — LangChain, CrewAI, Mastra, AutoGen — doesn't matter

GitHub热门项目 | Summer 2026 software engineering, data science, AI, quant, product management, and hardware internship postings. Updated daily by Simplify and Pitt CSC. | Stars: 45,065 | 18 stars today | 语言: Python

GitHub热门项目 | openpilot is an operating system for robotics. Currently, it upgrades the driver assistance system on 300+ supported cars. | Stars: 61,636 | 67 stars today | 语言: Python

I spent 13 years in financial services — 7 at Discover Financial, 6 at Bread Financial — consistently finishing in the top 5% of my team. I was good at my job. Really good. But in March 2026, I enrolled in Coding Temple's Full-Stack Web Development bootcamp and started building. Here's what 3 months actually looks like from zero. Month 1: HTML, CSS, and Figuring Out Why Nothing Looks Right I started where everyone starts — HTML and CSS. Built a food landing page (FoodSpot) and a multi-page event site (EventHive). Learned Flexbox, Grid, responsive design, and why box-sizing: border-box should just be the default everywhere. What I shipped: FoodSpot — food landing page EventHive — responsive multi-page event site What I earned: ✅ Web Development with HTML & CSS (Coding Temple verified badge) Month 2: JavaScript, Then Python JavaScript clicked faster than I expected. DOM manipulation, ES6+, event listeners. Then Python — and honestly, Python felt natural. The OOP concepts made sense immediately. What I shipped: Python CLI Task Manager — persistent task app with file storage, OOP, exception handling Defeat the Evil Wizard — text-based RPG with multiple classes, inheritance, combat logic, and game state management What I earned: ✅ JavaScript Mastery ✅ Python Foundations for Software Engineering ✅ Advanced Python Month 3: React React was the biggest jump. Component architecture, hooks, state management, routing. But I got through it by building something real. What I shipped: FakeStore API — a full e-commerce SPA consuming a live REST API with dynamic product rendering, client-side routing, CRUD operations, and loading/error state management What I earned: ✅ Single Page Apps with React What I Brought From Finance That Helped People underestimate what non-tech backgrounds bring to code. Here's what transferred directly: Data analysis → Debugging mindset. I spent years finding patterns in account data. Finding why code breaks is the same muscle. Process optimization → Clean

You just got your hands on an AI agent. It writes code, researches things, sends emails, books meetings. You feel like you're holding a chainsaw. But you keep using it to cut butter. The problem nobody talks about The gap between what your agent knows and what it can do is almost always a paywall, a KYC wall, or an API key. Here's what 'just add one data source' actually looks like: Go to the site. Click pricing. Choose a plan. Enter your email. Wait for verification. Click the link. Set a password. Enable 2FA. Download an authenticator app. Scan the QR code. Enter the 6-digit code. Fill in your company name. Add a credit card. Agree to terms. Find the API section. Generate a key. Copy it. Paste it into your code. Realize your agent doesn't know how to use it. Write a wrapper. Test it. Hit the rate limit. Add retry logic. That's one data source . Some workflows need ten. What x402 actually does Your agent hits an endpoint, gets a 402 (Payment Required) response with payment terms, pays a fraction of a cent in USDC or sats, gets the data back. No accounts. No API keys. No subscriptions. No puzzles. No humans in the loop. The concrete version Competitor research workflow: POST /company-info {"domain": "competitor.com"} -- $0.03 Returns: industry, HQ, headcount range, tech stack, social links POST /github-user {"username": "their-cto"} -- $0.002 Returns: repos, commit frequency, stars, languages, last active POST /dns-lookup {"domain": "competitor.com", "type": "MX"} -- $0.001 Returns: mail provider Full competitor profile: under $0.04. Under 3 seconds. Lead enrichment on 500 domains: under $20, done overnight, zero human hours. Setup (one system prompt line) Get a free key first (no wallet, no email): curl -X POST https://api.ideafactorylab.org/proxy/keygen Returns your key and an agent-ready prompt. Then tell your agent: You have a Cinderwright key. POST to https://api.ideafactorylab.org/proxy/do with header X-CW-Key and body {"task": "describe what you need in plain

We have tools for checking whether a query is injectable. We have linters, scanners, ORMs, parameterized queries, and database policies. But after the database returns rows, most applications simply trust that the result set matches the operation that asked for it. queryguard starts there. The query may be safe. The result may still be wrong. SQL injection taught us to distrust query construction. Parameterized queries answered the question: Did the user control the query structure? That question is well understood. The tooling is mature. But it is a different question from the one queryguard asks: Did this operation receive only the rows and fields it was allowed to receive? Those two questions are not the same. A perfectly safe parameterized query can still return the wrong row — because a predicate was dropped, a join widened the result, a developer selected a column they shouldn't have, or a query was rewritten without updating its scope contract. queryguard is not a database firewall. It is not a SQL injection scanner. It is not an ORM plugin. It is a contract check for observed result sets. Where it sits The hook position is the core design decision. queryguard sits immediately after cursor execution — before any result shaping, filtering, serialization, or response mapping. cursor = conn . execute ( sql , bindings ) rows = [ dict ( row ) for row in cursor . fetchall ()] evidence = queryguard . run_check ( contract , { " contract_id " : " user_profile_lookup " , " contract_version " : " 0.1.0 " , " params " : { " user_id " : user_id }, " session " : { " tenant_id " : tenant_id }, " result " : rows , }) if evidence [ " verdict " ] != " PASS " : raise QueryguardViolation ( evidence ) return rows Not at the HTTP layer. Not inside the ORM. Not at the API gateway. Immediately after the cursor returns rows — while the result is still raw, before anything shapes or discards it. This is intentional. If rows are shaped before queryguard sees them, queryguard cannot det

GitHub热门项目 | A list of free LLM inference resources accessible via API. | Stars: 24,157 | 100 stars today | 语言: Python

GitHub热门项目 | 小红书笔记 | 评论爬虫、抖音视频 | 评论爬虫、快手视频 | 评论爬虫、B 站视频 ｜ 评论爬虫、微博帖子 ｜ 评论爬虫、百度贴吧帖子 ｜ 百度贴吧评论回复爬虫 | 知乎问答文章｜评论爬虫 | Stars: 52,587 | 347 stars today | 语言: Python

GitHub热门项目 | Transforms complex documents like PDFs and Office docs into LLM-ready markdown/JSON for your Agentic workflows. | Stars: 69,150 | 524 stars today | 语言: Python

GitHub热门项目 | AI 时代的伯克希尔：基于 Claude Code 的价值投资研究框架。巴菲特·芒格·段永平·李录四大师方法论 + 多Agent并行研究。| AI-era Berkshire: a value investing research framework built on Claude Code. 4 masters' methodologies + multi-agent adversarial analysis. | Stars: 1,563 | 201 stars today | 语言: Python

AI harnesses are the new vendor lock-in. To swap across harnesses easily without rewriting your app, LiteLLM launched the Lite-Harness SDK . Run your prompt across different harnesses: from lite_harness import query , AgentOptions prompt = " Fix the failing test " # Claude Code harness async for message in query ( prompt = prompt , options = AgentOptions ( harness = " claude-code " , model = " claude-opus-4-8 " ), ): print ( message ) # Codex harness async for message in query ( prompt = prompt , options = AgentOptions ( harness = " codex " , model = " gpt-5.5 " ), ): print ( message ) To enable cost controls, fallbacks, and logging, point it to your LiteLLM AI Gateway: export LITELLM_API_BASE = https://litellm.your-company.com/v1 export LITELLM_API_KEY = sk-litellm-... Engineer's Takeaway: This SDK unifies how you invoke the agents, not how they run internally. Each harness keeps its native loop and tool-calling semantics. It is perfect for A/B testing agent performance and centralizing costs, but remember it is in public beta, so custom tool injection might require extra work! The Problem I Had My team was building an internal bot to fix failing CI/CD tests. We had three engineers advocating for three different harnesses: one wanted Claude Code, another Codex, and another Pi AI. Without an abstraction layer, we would have had to maintain three forks of the same bot , with three different SDKs, three logging systems, and three ways to track costs. It would have been an impossible maintenance burden. How Lite-Harness Helped The SDK solved that exact pain point in three concrete dimensions : 1. Unified Invocation (Time Savings) Instead of maintaining three separate implementations, I had a single query() that routed to whichever harness I wanted. Switching from Claude Code to Codex was literally just changing a string in the options. This allowed us to do real A/B testing in production for two weeks without rewriting any core logic. 2. Cost Observability (The Killer

Building a unified market intelligence platform for traders, analysts, researchers, and developers. After months of development, Jungletrade is now publicly available. The idea behind Jungletrade is simple: modern market analysis has become fragmented. Market data, indicators, analytical models, and trading signals are often distributed across multiple platforms, forcing users to maintain several subscriptions, workflows, and dashboards just to build a complete market view. We wanted to explore a different approach. 📊 The Problem Most market platforms focus on a specific layer of the analytical stack: Raw data Technical indicators Quantitative models Trading signals Each layer provides value, but users are frequently required to move between multiple tools to connect the pieces. Our goal was to create a modular ecosystem where these layers can coexist within a single platform. 🧭 The Jungletrade Ecosystem Today, JungleTrade provides four product categories: 📦 Data Structured datasets for market research and discovery. 🧠 Models Analytical frameworks designed to identify patterns and relationships within market data. 📈 Indicators Tools that transform raw information into actionable insights. ⚡ Triggers Event-driven signals designed to highlight potential market opportunities. 🔍 Built for Transparency One design decision was particularly important to us: every product should explain itself. Each product includes: Product description Key features Use cases Interpretation guidelines Methodology overview The objective is not simply to provide charts but to explain the problem being solved and how the underlying analysis works. 🔌 API First All products available through the platform are also accessible through API endpoints. Developers interested in integrating JungleTrade data into their own applications, dashboards, or research pipelines can request a demo API key through the platform. 🏗️ Architecture JungleTrade is built using a modular, service-oriented architecture des

I started as a Java developer, but for some time now, I have broadened my horizons. Recently, I thought about how early languages were dedicated to a single target and platform, and now they are broadening their focus. In this post, I want to write down my thoughts in the hope that it may be useful to others, probably to my future self. Definitions You may have been wondering about the title terms. I'm pretty sure that if you read this post, you have a pretty good picture of what a programming language is. Some may disagree on some finer points or raise a hair-splitting one, but it's not a PhD thesis, only a post on my blog. I must define what I mean by target and platform in the context of this post before going further. Target A target only makes sense in the context of compiled programming languages. For example, C's target is native code , and Java's is bytecode . Platform A platform is the system that will ultimately run the target. Native code runs on the operating system; bytecode on the JVM. Early programming languages Early programming languages had a single target and platform. I mentioned C and Java, but Ruby, Python, JavaScript, etc., were all the same. Programming language Target Platform C Native code Operating system C++ Native code Operating system Java Bytecode JVM Python - Python runtime TypeScript JavaScript Browser & server-side JS JavaScript - Browser I believe it was the case for a long time. It changed at some point, though. Multi-target is the new black The first time I heard about multi-target was in Scala. Scala came from the era of single-target and targeted bytecode on the JVM platform. However, in 2015, Martin Odersky announced Scala.js, which added JavaScript to Scala's target. The original article was published on InfoWorld, but it seems to have redirection issues nowadays. Here's the introduction on a copy: Scala, developed as a functional and object-oriented language for the JVM, is now multiplatform, with developers using it in abun