← Back to All Products
Experiments & Research · RPA AutomationResearch Prototype · Open Source
🤖

RPA Agent Skills System

Multi-Engine RPA Skills Platform for Anti-Bot, Reverse-Engineering & Mobile Automation

The RPA Agent Skills System abstracts browser, mobile, JS reverse-engineering, anti-bot, and workflow-orchestration capabilities into a standardized skill set (Skill + MCP dual-shell) that AI Agents can call directly. It integrates open-source engines such as CloakBrowser / Obscura / BrowserCluster / AgentGo / Lamda, hiding multi-engine fragmentation behind a unified interface, so Agents can treat "automation" as composable, reusable tools even against real-world scenarios like strong anti-bot, encrypted params, and mobile control.

Core Philosophy

Upgrade "RPA automation" from scattered scripts into a standardized skills platform that Agents can schedule.

🧩

Dual-Shell Same-Source

DUAL-SHELL

Business logic is written once (skills_impl); both the Claude Skill CLI and the single-process MCP are thin wrappers, avoiding dual maintenance and behavior drift.

🔌

Unified Engine Abstraction

UNIFIED ENGINE

Converge the heterogeneous APIs of CloakBrowser / Obscura / BrowserCluster / AgentGo / Lamda into one interface; Agents need not care about the underlying engine.

🌐

direct / cluster Dual-Mode

DUAL-MODE

Browser skills run locally against the engine (direct) or submit tasks to BrowserCluster for remote Worker execution (cluster) — handy in dev, scalable in prod.

🗂️

Skill-as-Tool

SKILL-AS-TOOL

Each capability is exposed as a namespaced tool (browser.* / proxy.* / data.* / antibot.*); Agents compose RPA flows like calling functions.

Industry Pain Points · RPA Skills System Answers

In real automation, the hardest part is never "click once" — it is bypassing defenses, reproducing encryption, and holding sessions.

🛡️

Strong Anti-Bot Blocking

Akamai BMP / Cloudflare Turnstile / reCAPTCHA v3 / DataDome / Kasada block naive scripts instantly. anti-bot-defender generates adversarial cookies, solves challenges, and reports fingerprint risks.

🔐

Hard Front-End JS Reverse

Request params are encrypted and signing algorithms hide in obfuscated JS, hard to reproduce stably. web-reverser systematizes via Observe→Hook→Breakpoint→Trace→Extract→Rebuild.

📱

High Mobile-Automation Barrier

Android devices are scattered, cloud-phones/emulators are complex to onboard, UI automation is fragile. mobile-controller offers 160+ APIs via Lamda for discovery, UI automation, remote desktop.

🍪

Sessions & Fingerprints Expire

Login state is hard to persist, cross-session reuse is hard, a fingerprint change triggers risk control. session-manager snapshots/restores Cookie & LocalStorage, with CloakBrowser persistent Profile.

🌍

Proxy Quality & Geo-Match

High ban rate, geo mismatch triggers risk control, slow responses hurt crawling. proxy-pool-manager offers 4 strategies: round-robin / least-failed / geo-match / fastest.

🔑

Credential & Sensitive Leak

Hardcoded keys, plaintext logs, scattered credentials hard to audit. security-guardian does static credential audit, SecureString validation, log masking, audit trail.

🧵

Multi-Step Flows Hard to Reuse

RPA flows are hand-wired, non-visual, hard to reuse. workflow-orchestrator orchestrates via DAG; task-scheduler takes over cron/interval execution.

🧩

Engine Fragmentation

Each open-source engine has a different API, high switching cost, capabilities can't compose. Engine Layer unified abstraction + Skill Layer composition make capabilities Lego-like.

Core Capabilities · 12 Skills

A skill matrix spanning browser, mobile, reverse-engineering, anti-bot, orchestration, and security, ready for Agents to orchestrate.

🌐

browser-controller

Browser lifecycle management, direct/cluster dual-mode: launch / navigate / click / type / screenshot / execute_script.

📊

data-extractor

Multi-mode structured extraction: XPath / CSS / GNE / LLM, covering news sites and SPA API interception.

🛡️

anti-bot-defender

Counter Akamai / Cloudflare / reCAPTCHA / DataDome / Kasada: generate adversarial cookies, solve challenges, report fingerprint risks.

🔓

web-reverser

JS reverse-engineering workflow: page observation, hook sampling, breakpoint debugging, network analysis, local rebuild.

📱

mobile-controller

Android automation: device discovery, UI automation, OCR, remote desktop — 160+ APIs via Lamda.

🧵

workflow-orchestrator

DAG visual flow orchestration, forwarded to BrowserCluster Vue Flow backend — no self-built DAG engine.

🌍

proxy-pool-manager

Proxy IP lifecycle management, 4 allocation strategies, SQLite persistence and stats.

🍪

session-manager

Multi-platform session snapshot/restore/import-export: Browser / Desktop / Mobile session types.

🔑

security-guardian

Static credential audit, SecureString validation, sensitive-data encryption and audit trail.

task-scheduler

Scheduled-task orchestration: Cron / fixed-interval / event / queue triggers, APScheduler-driven.

📚

site-knowledge-builder

Whole-site crawl → Markdown knowledge base (httpx + html2text), feeding downstream retrieval and RAG.

🛠️

skill-scaffold-generator

Generate the four-piece skill scaffold to quickly crystallize new Skill / MCP Server / Coded Agent.

Full Link of One Automation

From Agent intent to result return, the RPA Skills System chains defense, extraction, and orchestration into a reusable pipeline.

Engineering Flow
1
Intent Parsing & Skill Selection
Agent parses the task, picks browser / data / antibot from the skill matrix, and builds this run's execution plan.
2
Identity & Session Prep
proxy-pool-manager allocates a proxy; session-manager restores/creates a session; CloakBrowser loads persistent Profile and fingerprint.
3
Anti-Bot Defense
anti-bot-defender generates adversarial cookies / solves challenges; web-reverser reverse-engineers encrypted params when needed to bypass risk control.
4
Execute & Extract
browser-controller drives the page; data-extractor pulls structured data via XPath/CSS/GNE/LLM.
5
Orchestrate & Schedule
workflow-orchestrator chains multi-step flows; task-scheduler takes over timed/periodic tasks; cluster mode hands off to remote Workers.
6
Audit & Crystallize
security-guardian audits credentials and masks logs; skill-scaffold-generator crystallizes repeated patterns into new skills, closing the loop.
Defense First
Complete anti-bot and fingerprint prep before actual crawling, avoiding remediation after risk control triggers — sharply lowering ban rate.
Composable Capability
Each skill is a namespaced tool; Agents freely compose per task without rewriting integration layers for every scenario.
Loop Compound
Repeated flows are crystallized into reusable skills by skill-scaffold-generator; automation experience accumulates across projects.

Key Results · Engineering Asset

Upgrade "automation" from scattered scripts into an orchestratable, auditable, reusable skills platform.

12
Standardized Skills
9 implemented + Phase 4 placeholder
5
Unified Engines
CloakBrowser/Obscura/Cluster/AgentGo/Lamda
4
Proxy Strategies
Round-robin/Least-failed/Geo/Fastest
Dual-Shell
Skill + MCP
Same-source logic · zero drift

Key Breakthroughs

The difficulty is not single-point scripts, but engineering fragmented engines and defenses into a platform Agents can stably schedule.

🧩

Dual-Shell Same-Source

Business logic written once; Skill CLI and MCP process are thin shells, eliminating dual maintenance and behavior drift, minimal cost to add skills.

🔌

Unified Engine Abstraction

Converge CloakBrowser / Obscura / BrowserCluster / AgentGo / Lamda into one interface; Agents ignore the underlying differences and compose freely.

🛡️

Built-In Defense

Anti-bot, JS reverse, session/fingerprint, credential audit are first-class citizens of the skill matrix, ready out of the box.

♻️

Compounding Capability

workflow-orchestrator + skill-scaffold-generator crystallize repeated flows into reusable, parallelizable, generatable skill assets.

Business Value

Convert technical capability into quantifiable, reusable business value.

Improvements Delivered
Unify CloakBrowser / Obscura / BrowserCluster / AgentGo / Lamda behind one interface, letting Agents compose RPA capabilities like calling functions.
Promote anti-bot, JS reverse, session/fingerprint, and credential audit from add-ons to first-class citizens of the skill matrix — ready against real defenses.
direct/cluster dual-mode: local engine for handy dev, submit to BrowserCluster for production scale — capabilities migrate seamlessly.
Dual-shell same-source + skill scaffold make adding and crystallizing skills very cheap; automation experience compounds instead of being consumed once.
Applicable Scenarios
Competitor/partner site & business-flow research (within authorized scope): browser crawl + reverse + anti-bot combined.
Large-scale web structured crawling: data-extractor multi-mode + proxy-pool + session persistence.
Mobile Android automation & remote control: mobile-controller on Lamda 160+ APIs.
Orchestratable, schedulable, auditable RPA platform: workflow + scheduler + security combined in production.
📌

Content Under Active Update

Product content has been published based on internal materials. The following areas are planned for further development:

More real-site defense benchmarks and bypass-success-rate statistics
Remote scheduling & scaling metrics of Worker Agent in cluster mode
Integration with existing data platforms / warehouses and landing cases
Further refinement of compliance boundaries and authorized-scope statements
Contact Us

Start AI Partnership

Whether in government, finance, manufacturing, consumer, or content, we can customize vertical AI agent solutions for you.

📍

Address

Xiamen, Fujian · Wuhan OPC (planned)

🌐

Xianma AI

Xiamen Xianma Intelligent Technology Co., Ltd.

© 2024-2026 Xiamen Xianma Intelligent Technology Co., Ltd. · AI Agent Solutions · www.xianma.top

Products: 15active projects