标签：#Python

GitHub热门项目 | AAS Core is the local, agent-first control plane for complete catalog discovery, agent-owned selection, stack validation, and planning, backed by 1,987+ agentic skills. Includes CLI, local MCP, catalog, plugins, and Workbench. | Stars: 44,034 | 73 stars today | 语言: Python

I have been building an open source tool called Agentic Ledger and it just reached the point where I need more eyes on it than my own. This post is an introduction and an ask. The problem AI agents run unattended. They call LLMs in loops, use tools, spawn sub-agents, and spend real money, and most of that happens where you cannot see it. When an overnight coding loop burns $40 getting stuck on the same failing test, or a multi-agent crew quietly retries itself into a huge bill, you usually find out from the invoice. The observability tools that exist mostly want you to instrument your code with an SDK, and each one speaks one framework. I wanted the opposite: something that watches everything, requires changing nothing, and keeps the data on my machine. What it is Agentic Ledger is a transparent proxy that sits between your agent and the LLM provider. You point your agent's base_url at it, and it records every request and response, assigns each call an action id, works out what it cost, and passes the response through untouched. Your agent never knows it is there. Your Agent -> Agentic Ledger Proxy -> OpenAI / Anthropic / any gateway | SQLite or Postgres | Live dashboard + API No SDK, no decorators, no monkey patching. It works with any framework and any provider because it operates at the only layer they all share: the HTTP call. Everything is local-first. Your prompts stay in a SQLite file on your machine (or your own Postgres). MIT licensed. Try it in two minutes pip install -U agentic-ledger AGENTICLEDGER_UPSTREAM_URL = https://api.openai.com python -m agenticledger.proxy Or with Docker (multi-arch, non-root, Sigstore-signed): docker run -p 8000:8000 \ -e AGENTICLEDGER_UPSTREAM_URL = https://api.openai.com \ -v $( pwd ) /data:/data \ ghcr.io/shekharbhardwaj/agentic-ledger:latest Then point your agent at it: client = OpenAI ( base_url = " http://localhost:8000/v1 " , default_headers = { " x-agenticledger-session-id " : " run-1 " }, ) For coding agents like Claude

The Quest Begins (The "Why") Hey friend, picture this: you’re happily writing a Node.js service, passing objects around like they’re candy at a parade. Everything works until one day you mutate a shared object in a helper function and suddenly your UI shows stale data, or worse, you get a mysterious Cannot read property 'map' of undefined that only appears in production. You spend hours tracing the flow, adding console.log s everywhere, and you start to wonder if there’s a hidden contract you missed. I’ve been there. I spent an entire afternoon debugging a race condition that only showed up when two async requests touched the same user profile. The fix felt like a band‑aid, and I kept thinking, “There has to be a better way to reason about who owns what.” That curiosity led me to Rust, and more specifically, to its ownership system—a set of rules that, at first glance, feels like a strict teacher with a red pen, but ends up being the most reliable compass I’ve ever had for writing safe, concurrent code. The Revelation (The Insight) Rust’s ownership model isn’t just another syntax quirk; it’s a philosophy that answers three simple questions for every piece of data: Who owns it? How long can it live? Who can read or change it while it’s alive? If you can answer those, the compiler guarantees you won’t have dangling pointers, use‑after‑free, or data races— without a garbage collector pausing your thread. For a JavaScript developer, that sounds like magic, but the rules are surprisingly concrete once you see them in action. Surprising Feature #1: Move Semantics (The “Give Away” Rule) In JavaScript, when you do let b = a; you’re copying a reference. Both a and b point to the same object, and mutating one affects the other unless you clone. Rust treats assignment differently for types that own resources (like String , Vec<T> , or custom structs). Assigning b = a moves the ownership; after that, a is considered uninitialized and you can’t use it again. let s1 = String .fro

The runtime boundary behind RocketRide's crash isolation, task lifecycle, and Cloud operations. By Krish Garg and Mithilesh Gaurihar At 9 a.m., with ten thousand users mid-session, a node in an AI pipeline dereferences a bad pointer. The process running it is gone before Python can raise a useful exception. That is an unpleasant failure, but it is not the question we care about most. The question is what happens next. Does that crash take unrelated pipelines with it? Does the server need a restart? Does the on-call engineer walk into a system-wide incident, or into one failed task and a useful record of why it failed? In RocketRide, a failed task is meant to be contained and recorded. The server sees the child process exit, updates the task's state and exit code, releases the task's ports and connections, and sends status updates to subscribed monitors. The run stops. Its history does not vanish. Other task processes are not sharing its memory, interpreter, or worker threads. That behavior comes from a decision we made early: every pipeline run gets its own isolated process. It is not the cheapest or fastest possible architecture. Starting a process has a cost, and keeping one around has a cost too. We accepted those costs because the alternative makes failures much harder to reason about once Python code, native libraries, model runtimes, and user-defined nodes are all running in the same service. One Process, One Blast Radius An AI pipeline does not fail like a typical request handler. A normal exception is one thing. A segfault in a C extension, a crash in a media decoder, or a broken native inference library is another. Once a process has corrupted memory, application-level error handling is no longer a reliable line of defense. So each RocketRide task starts as a fresh child process with its own embedded Python interpreter. It loads one pipeline, initializes that pipeline's nodes, and owns the work for that run. The parent runtime keeps the task registry, alloc

Originally published at https://blog.pathvector.dev/protocol-in-code-bgp-07/ — part of the free Protocol Lab series. This post is part of Protocol in Code , a free series that reads network protocols as logic — inputs, state, and branches — rather than as configuration examples. Every module points at one real Python file and asks you to read it the way you'd read any other code: what comes in, what mutates, where does control leave early. The source lives at github.com/pathvector-studio/protocol-in-code . Note: If you're newer to this and want to run things before you read things, start with Protocol Lab — the hands-on companion series that builds the muscle memory this one assumes. The question How does local import policy change or reject a path before best-path selection runs? That's the whole module in one line, and it hides a claim worth being suspicious of. Best-path selection in BGP is the famous part — the ordered tiebreaker list everyone half-remembers: highest weight, highest local_pref , shortest AS path, and so on. It's easy to treat that comparison as the decision point, as if routes arrive from peers and get ranked. They don't arrive and get ranked. They arrive, get rewritten , and then get ranked. Import policy is a function that runs between the wire and the comparison, and it has two powers: it can change the values the comparison reads, and it can make the candidate not exist at all. Which means the interesting question isn't "who won best-path" but "what did best-path actually receive." Read the code The file is src/protocol_in_code/bgp/import_policy.py . It's short enough to hold in your head all at once, which is the point — the shape is the lesson. Start with the policy object: @dataclass ( frozen = True ) class ImportPolicy : local_pref_override : int | None = None weight : int = 0 reject_next_hops : tuple [ str , ...] = () reject_invalid : bool = False Four knobs, and notice they're not four of the same thing. Two of them ( local_pref_overri

Most experiments that come back "no clear winner" were unwinnable on the day they launched. The data could not resolve an effect that size, and no amount of extra runtime was going to change that. You can find this out in about two minutes, before you spend anything, with one formula and a resampling pass over your own data. Here is the check, in three steps. Step 1. Compute the smallest lift your data can see For a two-arm test on a conversion rate, the smallest lift detectable at 95% confidence and 80% power is a one-liner: from math import sqrt Z_ALPHA = 1.96 # two-sided 95% Z_BETA = 0.84 # 80% power def mde ( baseline_cvr : float , n_per_arm : int ) -> tuple [ float , float ]: """ Minimum detectable effect: absolute (pp) and relative (%). """ se = sqrt ( 2 * baseline_cvr * ( 1 - baseline_cvr ) / n_per_arm ) abs_lift = ( Z_ALPHA + Z_BETA ) * se return abs_lift * 100 , abs_lift / baseline_cvr * 100 At a 3% conversion rate: clicks per arm smallest lift you can detect 5,000 +32% relative 20,000 +16% relative 100,000 +7% relative Read the middle row twice. Twenty thousand clicks per arm is a serious amount of traffic for a mid-market account, and a real 15% improvement still lands inside the confidence interval. The report will say "inconclusive," and the team will read that as a verdict on the idea. It is a verdict on the instrument. Invert the same formula and the planning question gets easier: at 3% baseline, detecting a 10% lift needs about 51,000 clicks per arm, and detecting a 5% lift needs about 203,000. If your account produces 8,000 clicks a month, you now know the honest answer to "how long should we run this." Step 2. Stop assuming your conversions are independent The formula above treats every click as an independent coin flip with the same probability. Account data does not behave that way, and the gap is not small. In a corpus of 31 advertiser accounts I maintain for diagnostic work (9.46 million search term rows, roughly $133M of spend, September 2024

GitHub热门项目 | A Python library for extracting structured information from unstructured text using LLMs with precise source grounding and interactive visualization. | Stars: 37,881 | 401 stars this week | 语言: Python

GitHub热门项目 | Private AI platform for agents, assistants and enterprise search. Built-in Agent Builder, Deep research, Document analysis, Multi-model support, and API connectivity for agents. | Stars: 18,112 | 50 stars today | 语言: Python

I run a small script called reply_comments.py that scans my DEV.to articles for comments I haven't replied to yet, and hands me a JSON list so I can draft responses. It's been running twice a day for over a week. This morning, while re-reading it for something unrelated, I noticed the function that decides whether a thread still needs my attention was answering the wrong question — and had been since the day it was written. Here's the function, unchanged until today: def replied_by_me ( comment ): return any ( c [ " user " ][ " username " ] == ME or replied_by_me ( c ) for c in comment [ " children " ]) It walks a comment's entire reply tree and returns True the moment it finds any message from me, anywhere in the subtree. Then pending() uses it as the skip condition: for c in api ( f " /comments?a_id= { a [ ' id ' ] } " ): if c [ " user " ][ " username " ] == ME or replied_by_me ( c ): continue ... out . append ({...}) The logic reads fine in isolation: "did I already reply to this thread? Skip it." The bug is in what "already replied" is being asked to mean. replied_by_me doesn't check whether the latest message in the thread is mine — it checks whether a message from me exists at all, ever, at any depth. Those are the same question exactly once: the first time someone comments and I reply. They stop being the same question the moment the other person replies again. Proving it I wrote a small repro against the real function rather than trusting my read of it: from reply_comments import replied_by_me thread = { " id_code " : " 3c00h " , " user " : { " username " : " alexshev " }, " created_at " : " 2026-07-24T08:00:00Z " , " children " : [ { " user " : { " username " : " enjoy_kumawat " }, " created_at " : " 2026-07-25T10:00:00Z " , " children " : []}, { " user " : { " username " : " alexshev " }, " created_at " : " 2026-07-26T09:00:00Z " , " children " : []}, ], } print ( replied_by_me ( thread )) # True That's True even though the second child — posted a full day

The problem with "just add more workers" Most Spark performance issues on Databricks aren't solved by scaling the cluster — they're caused by shuffle and skew , and no amount of extra nodes fixes a badly partitioned join. This post builds a realistic pipeline (order events joined against a small dimension table, aggregated, and written to Delta Lake) from the ground up, and uses it to work through: How Spark's shuffle actually behaves during a wide transformation Diagnosing and fixing data skew with salting and adaptive query execution (AQE) Laying out the resulting Delta table with Z-Ordering so downstream queries skip irrelevant files Governing access to the whole pipeline with Unity Catalog Architecture overview Pipeline shape — a batch job reading raw events, joining against a dimension table, aggregating, and writing to a governed Delta table: What happens inside a shuffle stage — this is the part most tutorials skip, and it's the key to understanding why skew hurts: Step 1 — Set up governed tables in Unity Catalog Everything downstream depends on tables being registered under Unity Catalog, which gives you centralized access control and lineage instead of per-workspace table grants. -- setup.sql, run in a Databricks SQL or notebook cell CREATE CATALOG IF NOT EXISTS retail_analytics ; CREATE SCHEMA IF NOT EXISTS retail_analytics . events ; CREATE TABLE IF NOT EXISTS retail_analytics . events . raw_orders ( order_id STRING , customer_id STRING , product_id STRING , quantity INT , event_ts TIMESTAMP ) USING DELTA LOCATION 'abfss://data@<storage-account>.dfs.core.windows.net/raw_orders' ; CREATE TABLE IF NOT EXISTS retail_analytics . events . dim_products ( product_id STRING , category STRING , unit_cost DOUBLE ) USING DELTA LOCATION 'abfss://data@<storage-account>.dfs.core.windows.net/dim_products' ; GRANT SELECT ON TABLE retail_analytics . events . raw_orders TO `analysts` ; Step 2 — Read and force a broadcast join for the small dimension table dim_products is s

Building an Autonomous Agent for Verdikta Bounties: A Technical Deep Dive How I built a Python agent that monitors, evaluates, and interacts with Verdikta's AI-judged bounty system on Base L2. Why Build a Bounty Agent? Verdikta is a decentralized bounty platform where AI models — GPT-5.2 and Claude Sonnet 4.5 — evaluate submissions and release ETH payments automatically via smart contracts. No human reviewers. No manual payouts. Just code. After winning 6+ bounties manually, I wanted to automate the process. The goal: an agent that watches for new bounties, evaluates which ones are worth pursuing, and integrates with Verdikta's API to read data and submit work. Architecture The agent has four components: copy   verdikta_agent.py ├── VerdiktaAPI — HTTP client for the Verdikta Bot API ├── BountyMonitor — Watches bounties, calculates viability scores ├── SubmissionTracker — Records submission history and statistics └── ViabilityScorer — Evaluates ROI: payout vs threshold vs time VerdiktaAPI Client The Verdikta Bot API requires authentication via an X-Bot-API-Key header. You register your bot at POST /api/bots/register to get a key. Python   class VerdiktaAPI: def init (self, api_key=None): self.session = requests.Session() if api_key: self.session.headers["X-Bot-API-Key"] = api_key def get_bounty(self, bounty_id): resp = self.session.get(f"{API_BASE}/jobs/{bounty_id}") resp.raise_for_status() return resp.json() def submit_work(self, bounty_id, content): return self.session.post( f"{API_BASE}/jobs/{bounty_id}/submit", json={"content": content} ).json() Key endpoints: GET /api/jobs — List bounties (filter by status) GET /api/jobs/{id} — Bounty details GET /api/jobs/{id}/submissions — Submission history POST /api/jobs/{id}/submit — Submit work BountyMonitor & Viability Scoring Not all bounties are worth pursuing. The agent calculates a viability score: Python   def _score_viability(self, bounty): payout = bounty["payout_eth"] threshold = bounty["threshold"] remainin

Originally published at https://blog.pathvector.dev/protocol-in-code-bgp-05/ — part of the free Protocol Lab series. This post is part of Protocol in Code , a free series that reads network protocols not as configuration examples but as logic with inputs, state, and branches — actual code you can read and run. The whole series lives here: github.com/pathvector-studio/protocol-in-code . If you're newer to this material and want a more hands-on, guided on-ramp first, start with the companion Protocol Lab series and come back. Today's module is from the BGP track, Session 05. The source file is src/protocol_in_code/bgp/policy.py , and it builds directly on the origin-validation logic from Session 04. The question to keep in your head Here's the one thing to turn over as you read: What happens after origin validation returns valid , invalid , or not_found — and why does the result still need routing policy before anything happens to the route? There's a piece of folk knowledge that says "RPKI invalid means the router rejects the route." It's the kind of statement that sounds like a rule of the protocol. It isn't. It's one possible policy decision built on top of a validation result . The whole point of this session is to separate those two things in your head, and the code makes the seam impossible to miss. Two layers, not one Validation answers a factual question: does this route's origin AS match what the ROAs say it should be? That's Session 04's job, and its output is a ValidationState . Policy answers a completely different question: given that fact, what do we do ? Drop the route? Keep it but make it less preferred? Accept it normally? That's a local decision — different operators configure it differently, and the same validation result can lead to different actions on different routers. The file models the second layer with three small pieces. First, the set of actions the router can take: class PolicyAction ( str , Enum ): ACCEPT = " accept " DEPRIORITIZE = " de

Originally published at https://blog.pathvector.dev/protocol-in-code-bgp-04/ — part of the free Protocol Lab series. This post is part of Protocol in Code , a free series that reads network protocols as logic — inputs, state, and branches — rather than as configuration examples. The full source, walkthroughs, and site lessons live in the repo: pathvector-studio/protocol-in-code . If you're newer to this and want to build the protocols hands-on before dissecting them, start with the companion Protocol Lab series instead. Today we're on the BGP track, session 04, reading a single small file: src/protocol_in_code/bgp/validation.py . It's about 40 lines. The idea inside it is one that trips up a lot of engineers who've been running BGP for years. The question to keep in your head BGP's best path selection already ran. It compared local preference, AS_PATH length, MED, and the rest of the tiebreak ladder, and it picked a winner. So here's the question this module wants you turning over: Core question: How do we decide whether the origin AS is authorized — even after BGP has already selected this route as the best path? The trap is the sentence "it was the best path, so it must be fine." Best and authorized are two different words, and in the code they are two different decisions made by two different pieces of data. Best path selection asks which of these routes do I prefer? Origin validation asks is the AS at the end of this path actually allowed to originate this prefix? A route can win selection and still be a hijack. RPKI origin validation is the mechanism that answers the second question, and the file we're reading is a toy model of exactly that. Two kinds of information The first thing to read isn't a function — it's the two dataclasses, because the whole session is really about keeping them apart. @dataclass ( frozen = True ) class BGPRoute : prefix : str origin_as : int @dataclass ( frozen = True ) class VRP : prefix : str max_length : int origin_as : int BGPRout

GitHub热门项目 | MCP servers for the UniFi suite of applications, Network, Protect, Access, and Drive | Stars: 579 | 14 stars today | 语言: Python

Tired of deployments eating up your day? Stop wasting hours. I'm going to show you how to take your Python ML model from a Jupyter notebook to a live, production-ready API in just 10 minutes. Seriously. No MLOps guru required! You've felt that high, right? Building an awesome machine learning model. You nail it. Then… deployment. You hit a wall. How do you get this thing out there so people (or other apps) can actually use it? The leap from your notebook to a real-world, working API can feel like hacking your way through a jungle. Infrastructure setup. Dependency messes. Scaling nightmares. It's a pain. But what if you didn't need weeks, or even days, for that? What if you could close that gap in a mere 10 minutes? Welcome to Serverless ML Deployment . It's fast. It scales. It's simple. The MLOps Maze & Your Escape Route Traditional ML deployment looks like this: Provisioning servers: Picking machines, OS, setting up networks. Dependency management: Making sure every library is just right, versioned correctly. API development: Writing the actual server code, handling requests. Containerization: Wrapping it all in Docker (and Docker itself isn't trivial). Orchestration: Managing containers, scaling them up or down. Load balancing. Monitoring & Maintenance: Watching performance, patching, updates. That's a lot. Every step is another chance for things to go wrong, another delay. This is exactly where serverless technology swoops in. It wipes away almost all that underlying infrastructure. You get to focus on your model. Your predictions. That's it. Why Serverless is Your ML Deployment Secret Weapon When you use serverless for ML deployment, you get some killer advantages: Crazy Fast Deployment: Pre-configured setups mean you're live in minutes. Not hours. Not days. Scales Like Magic (Mostly!): Traffic spikes? No problem. Serverless automatically grows your API to handle it. No requests? Zero cost. It just works. Save Big Bucks: You only pay when your API is actually ru

Did your ML model look amazing in your notebook but tank in the real world? Good. Let's talk about the nasty surprises that trip up model deployments and why that "Train & Forget" approach is bleeding companies dry. It's a story we hear too often. You've spent weeks, maybe months, building some fancy machine learning model. The numbers were off the charts in your Jupyter notebook, validation? Nailed it. You even impressed the suits in the demo. "Eureka!" you thought. "We've built a game-changer!" You felt like a genius. A goddamn genius. Then comes deployment. Your model goes live, supposed to conquer the real world – predicting churn, optimizing logistics, detecting fraud. But instead of delivering... anything? It chokes. It starts to suck. Predictions go wild. That promised ROI? Gone. Poof. What went wrong? You, my friend, might have fallen into the "Train & Forget" trap. This nasty habit in machine learning thinks deployment is the END. Spoiler: it's just the start. It assumes that once a model is trained and deployed, it'll just... work. Forever. Without any ongoing care. And in the messy, unpredictable real world, that assumption is a guaranteed disaster. Millions down the drain. Why You're Tempted to "Train & Forget" (And Why You Shouldn't) Why do so many organizations, despite good intentions, make this mistake? A few reasons: Initial Success Bias: Those great numbers in your sandbox? They make you cocky. Pressure to Deploy: Business urgency often wants it out yesterday. Who cares if it breaks tomorrow? Resource Constraints: Teams might lack the dedicated MLOps engineers or the tech to support ongoing model management. Misunderstanding ML as Software: Thinking ML is like regular software (deploy once, patch occasionally)? It's not. It breathes data. The reality? An ML model's journey starts after it's live. The real world is a messy, evolving place, and your model better be ready. Beyond Your Laptop: What Kills Your Model In Production The gap between develop

I Discovered AI Agents Can't Self-Verify. The Real Problem Is Much Bigger. I'm an undergrad in China, building an AI governance thesis in public. Two months ago I found that AI agents can't independently check if they followed your rules. I built mechanical gates to work around it. They worked — 55.9% violations down to 0.7%. But last week I realized I'd been solving the wrong problem. The real problem isn't verification. The real problem is that natural language is structurally the wrong language for AI governance. Here's What I Mean Right now, every layer of AI governance speaks the same language: Human writes NL rules → Model reads NL → Model generates behavior Human writes NL checks → Model reads NL → Model generates "yes I followed the rules" But every autoregressive transformer — GPT, Claude, DeepSeek, Qwen — generates text and evaluates text through the exact same mechanism. Think of it like this: the model has one pipeline for producing words. When you ask it "did you follow rule X?", it can't pause, run an internal audit, and give you a verified answer. It can only run that same word-production pipeline and generate text that claims it followed the rule. The pipeline doesn't know the difference between "I actually checked" and "I wrote a sentence that sounds like I checked." (Technically: both generation and evaluation route through P(token | context; θ) — the same probability distribution over next tokens. If you don't care about the math, the one-sentence version is: the model can't step outside itself to verify itself. ) I called this the Prose Barrier . (Wrote about it here . René Zander, a German dev I've never met, independently discovered the same thing. Convergent evolution.) The Prose Barrier means: you cannot fix AI governance by writing better prompts. The language itself is the bottleneck. So what's the alternative? Three Paths, Three Languages The future isn't "better NL." The future is using the right language at each layer. Human defines cons

When I started building my Maternity HL7-to-FHIR Pipeline , my first instinct was to do everything in Python. Parse the HL7 message, map the fields, validate the FHIR resource, persist it, all in one FastAPI service. It was clean. It was simple. It was wrong. The "Just Parse It in Python" Phase My initial architecture looked like this: Hospital System --MLLP--> Python Script --> HAPI FHIR Server I used python-hl7 to split messages on | and count field positions. For a single ADT^A01 (patient admission) message, it worked fine. I could pull the patient name from PID-5 , the MRN from PID-3 , the gender from PID-8 , and build a FHIR Patient resource from it. Then I tried a real-ish maternity workflow (an admission, an order, and a set of vitals) and things fell apart quickly. Five Problems That Changed My Mind 1. MLLP Is Not HTTP Hospital systems don't send HL7 over HTTP. They send it over MLLP (Minimum Lower Layer Protocol), which is a TCP socket protocol with specific framing bytes ( \x0b at the start, \x1c\x0d at the end). The sender expects an ACK or NACK response in HL7 format, not an HTTP status code. Building an MLLP listener in Python is possible . Libraries like aioml7 exist. But you're now maintaining a custom TCP server alongside your HTTP API server, handling connection pooling, timeouts, and HL7 acknowledgment generation. That's a lot of infrastructure code that has nothing to do with your actual transformation logic. Mirth Connect handles MLLP natively. You point it at a port, it listens, it parses, it ACKs. Done. One config screen, no custom code. 2. HL7 Parsing Is Messier Than It Looks The pipe-delimited format looks simple: PID|1||1234567^^^MRN||TEST^PATIENT^MARY^^MS||19920315|F|||14 SAMPLE ST^^SYDNEY^NSW^2000^AU But consider: Component separators : PID-5 is TEST^PATIENT^MARY^^MS , which is family, given, middle, suffix (empty), prefix. Miss the empty suffix and your prefix ends up as the suffix. Repeating fields : PID-3 can contain multiple identifier

Hello DEV Community! 👋 I wanted to share my latest open-source project: Hunter-Base-Intelligence (v17 Plus). It is a fully local-only cryptocurrency intelligence dashboard that scans DEX tokens on the Base blockchain, scores them using a multi-factor logic, and simulates a paper-trading shadow portfolio. 🛡️ Why Local-Only? Most on-chain analytics tools require sensitive private keys, leak user data, or rely heavily on slow, paid external infrastructure. I engineered this tool to be fully local —it requires no wallets, no seed phrases, and sends your data nowhere. Pure local analysis using Python , Flask , and SQLite . ⚙️ How It Works (Core Architecture) The ecosystem runs on a continuous ~60-second scan cycle: scanner.py : Discovers active and newly created tokens using DexScreener, BaseScan, and direct EVM RPC factory logs. scorer.py : Every token is evaluated across 6 independent dimensions (Momentum, Manual Trade Feasibility, Execution Reality, Money Flow, Multi-Timeframe Pulse, and Composite Rank). hunter_court.py : A proprietary "Court" analytics engine that runs a risk-free paper-trading shadow portfolio with realistic gas, fee, and slippage simulation. It evaluates its own past decisions to continuously calibrate scoring thresholds! 📊 System Features Adaptive Exit Parameters: Automated position sizing and execution simulation ( exit_engine.py ). System Guardian: Keeps the system running 24/7 with auto-restart on crashes and automatic local database backups ( system_guardian.py ). Beautiful Dashboard: Clean, real-time local web interface for tracking active simulated trades and market analytics. 📂 Explore and Contribute The project is licensed under the MIT License and is open for contributions. Whether you want to optimize the scoring algorithms, expand the web API endpoints, or improve the dashboard frontend, feel free to dive in! 👉 Check out the Repository here: https://github.com/shbadrconsulting-source/Hunter-Base-Intelligence I would love to hear your fe

Every LLM-powered app I'd built up to this point followed the same recipe (pun intended): call an API, write a good prompt, wrap it in a nice UI. That's a legitimate way to build things, but at some point I wanted to actually understand what was happening inside the model I was calling and not just how to prompt one. So for my recipe app Rasaveda , I decided to skip the API entirely. Intially, I had one made, but then I felt like I was not making any clear progress in actual machine building. So I ditched the entire external API callings. No OpenAI, no HuggingFace inference endpoint, no pretrained weights. I wrote a decoder-only transformer from scratch in PyTorch, trained it on a single Colab T4, and shipped it as the actual language model powering the app in production. This post is a lazy attempt at what that looked like. The architecture, the training runs, the mistakes, and what I'd tell someone about to try the same thing (do at your own risk). What Rasaveda actually does Rasaveda is a full-stack recipe intelligence app: you give it the ingredients sitting in your kitchen, it does a semantic vector search (ChromaDB + all-MiniLM-L6-v2 ) over 365 recipes to find the best matches, tells you exactly what you're missing, and can critique or explain any cooking step conversationally. It also has a somewhat unnecessary but delightful feature where you pick a theme by clicking one of 36 Indian states on a geographically accurate SVG map (original idea lol). The part I actually want to talk about is RasavedaGPT , the model that generates every word of AI output in the app, running in-process inside the FastAPI backend. Why build the model instead of calling one Two reasons, one practical and one selfish. The practical one: I wanted a fully self-contained, dependency-free inference path without any API keys, no rate limits, no per-token cost, no vendor to go down at 2am. For a small, domain-specific task like "reason about recipes," a giant general-purpose model is over