Skip to content
View ApexAiOfficial's full-sized avatar

Block or report ApexAiOfficial

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ApexAiOfficial/README.md

ApexAi: web data, business communications, and autonomous operations. Operators, software through ApexAPI, and AI agents through ApexMCP reach Scrawl, Communications, and DevOps, each through its own authority gate; DevOps access through ApexAPI and ApexMCP is planned. Opening for business soon.

Visit apexaiofficial.com Follow launch updates on Telegram Request a custom build

ApexAi builds guarded products for web data, business communications, and autonomous operations, with operator, API, and agent interfaces. Scrawl turns web sources into owned artifacts and versioned datasets, with crawling built on Scrapy. Communications runs customer support and business messaging on your own Twilio, SendGrid, and Telegram accounts, with consent and history built in and the AI provider of your choice, or none. DevOps corrects failures in the systems a customer connects, using only actions the customer allows, and proves each fix by observing the system again; its engine is in development and the managed service is planned.

Each product owns its rules. A person working in Telegram, an application calling ApexAPI, and an AI agent using ApexMCP all pass the same product policy before anything runs, and all get evidence back. DevOps will be reached the same way, but its ApexAPI and ApexMCP access is planned, not available yet.

Scrawl, web-data operations: turn web sources into owned artifacts and versioned datasets. Workflow from source to dataset: choose a cataloged source, pass the policy and egress gate, crawl, fetch, or render, classify and store artifacts, publish a dataset version, then search and fetch records.

Communications, customer support and business messaging: customer conversations on your own provider accounts, with consent, history, and your choice of AI. Workflow for a support conversation: the customer writes in, the conversation opens or reopens, a person, automation, or allowed AI answers, the consent and Trust and Safety gate, the reply goes out from the number they wrote to, then the reply and its provenance are kept in history.

DevOps, autonomous operations: find what broke, fix it within the authority you grant, and prove the fix held. Workflow from incident to verified resolution: a declared condition fails, one incident is opened, an allowed remedy is selected, a single-use grant is issued, the action runs within its bound, a fresh verification, the stability window holds, then it is closed with evidence. Engine in development; managed service planned.

Public availability. Scrawl: Telegram operator, ApexAPI, ApexMCP, and Scrawl Plugin are in development. Communications: Telegram operator, web console, and ApexAPI are in development; ApexMCP and the Communications / Twilio Plugin are planned. DevOps: the operations engine is in development; ApexAPI and ApexMCP access are planned. No product is publicly hosted yet.

Access layers. ApexAPI: typed HTTP for applications. ApexMCP: tools and resources for AI agents. Product plugins: operator control with approvals and evidence. Each product owns its contracts, and there is no shared gateway; DevOps access through ApexAPI and ApexMCP is planned.

Custom orders: bespoke software, scoped before it is built. Telegram and messaging, communications automation, web data, integrations and APIs, agents and AI tooling, and operations and control.

Open source from ApexAi: guides, kits, and tools you can use today.

privacy-chain: a guide to how online identity leaks across devices, networks, browsers, and accounts, and the controls that reduce it. telegram-bot-tutorial: build a Telegram bot in Python from zero to deployed. termux-kit: turn Android Termux into a practical workbench. termux-telegram-bot: deploy and operate one Python Telegram bot on Android Termux. cookie-manager: inspect, export, and replay browser cookies you control. All ApexAi public repositories

ApexAi: web-data, business communications, and autonomous operations products for operators, software, and AI agents.

apexaiofficial.com
Sales: sales@apexaiofficial.com
Support: support@apexaiofficial.com
On Telegram: official updates · community · Scrawl · Communications

Popular repositories Loading

  1. ytdlnis ytdlnis Public

    Forked from deniscerri/ytdlnis

    Unmodified fork of deniscerri/ytdlnis. No changes are published here; see upstream for the current app.

    Kotlin

  2. git-sms git-sms Public

    Forked from github/git-sms

    Contribution fork of github/git-sms. Branch revival/pr-1-restore-prototype restores the SMS prototype with regression tests and CI, proposed upstream in github/git-sms#1.

    Python

  3. ua-rotator ua-rotator Public

    Framework-agnostic weighted User-Agent rotation with an optional Scrapy adapter

    Python

  4. DeepSeek-Reasonix DeepSeek-Reasonix Public

    Forked from esengine/DeepSeek-Reasonix

    Fork of esengine/DeepSeek-Reasonix (main-v2) with one change: sha256sum -c manifest checks now count as delivery verification in the evidence gate. Otherwise identical to upstream main-v2.

    Go

  5. termux-kit termux-kit Public

    Practical Termux foundation for Android: local services, dev stack, helper scripts, and safe backups

    Shell

  6. termux-telegram-bot termux-telegram-bot Public

    Deploy and operate a single Python Telegram bot on Android Termux with strict token safety

    Shell