Skip to content

Aimeck15 v1.3.0

Latest
Latest

Choose a tag to compare

View all tags
@ashguine-svg ashguine-svg released this 08 Jun 14:18
v1.3.0
0971357
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Aimeck15 v1.3 Release Notes

🛡️ Important Notice Regarding Security and Safety (Please Read)

Aimeck15 is an advanced RPA tool that utilizes image recognition and automated mouse/keyboard control. Due to the technical nature of these operations, it may occasionally be falsely flagged (False Positive) as a "suspicious program" by certain security software, such as Windows Defender, or web browsers.

To ensure our users can utilize the software with complete peace of mind, we want to explicitly clarify the safety and transparency of this software:

  • 100% Local Processing: All image recognition and OCR processes in Aimeck15 are completed entirely on your local PC. Your screen capture data, entered passwords, and text information are never transmitted to any external servers.
  • Clean Security Tests: For this release, Aimeck15 has passed strict security evaluations to ensure safety:
    • VirusTotal Scan: Zero detections for malicious malware or viruses.
    • Virtual Machine (VM) Testing: Verified in a sandbox environment to ensure no suspicious background communications or spyware-like behaviors occur.
  • Developer Information: This software is designed and built by a Japanese developer, with software safety and user transparency as the absolute top priorities.

🚀 v1.3 Major Update

16. Multi-Action Editor

The Multi-Action Editor is a powerful feature used when you want to execute not just a single click, but multiple connected steps, such as "keyboard operations" or "text input," linked to a specific image trigger[cite: 1].

【Four Key Features of the Multi-Action Editor】

  • 1. Complex Conditional Branching with a Single Image: Easily handles complex "conditional branching" that was difficult to create using traditional single-image recognition alone[cite: 1]. By assigning multiple steps to a single image and promptly returning to the image recognition standby state upon operation completion, it effectively prevents automation stuck/freeze issues midway[cite: 1].
  • 2. Minimizing Operations with "Image Recognition Triggers": Instead of actively forcing operations like traditional RPA tools, it initiates actions "only when the target image (screen) is found."[cite: 1]. This drastically reduces false operations and unintended actions on unexpected screens or when the application is not launched[cite: 1].
  • 3. Visual and Smart Recording & Editing: It filters out scattered mouse movements during recording and extracts only the necessary actions[cite: 1]. "Pins" indicating the operation order and type are placed on the actual screen in real-time, making the recording state clear at a glance[cite: 1]. Additionally, the wait time (standby time) between operations can be batch-adjusted later at any time[cite: 1].
  • 4. High Maintainability via Flexible Scene Splitting & Merging: Even during long step recordings, it automatically splits scenes by detecting screen transitions[cite: 1]. You can freely split or merge operations later, such as "linking only a specific scene to a separate image" or "consolidating all operations into a single image."[cite: 1]. Post-recording adjustments can be made intuitively[cite: 1].

💡 Equipped with Proprietary "Natural Mouse" Functionality

  • By sampling human-specific hand movement habits, it eliminates the abrupt "cursor warping" typical of macros[cite: 1].
  • This achieves extremely natural and safe mouse trajectories that make it highly resilient against tool detection[cite: 1].

🛠️ Seamless Workflow for Recording, Editing, and Execution

  • Intuitive Recording Procedure: Start recording by pressing the [Record] button, select the target window, and enter a scene name[cite: 1]. Perform the actions you want to record, and Aimeck15 will track your actions while saving automatic screen split points (Auto-Split)[cite: 1]. Pins are automatically dropped on coordinates when operating on the app window[cite: 1]. The on-screen remote panel allows you to pause, undo actions, input text, and stop the recording[cite: 1].
  • Advanced Wait Settings: Using the "Auto Wait" setup config block, you can establish an automated batch delay time inserted across all action intervals, or manually incorporate a standalone "Wait" event block[cite: 1]. The delay timing defaults to 1.0s[cite: 1]. Cranking this interval down excessively will trigger empty click misses on heavy rendering software ecosystems, because subsequent automated actions fire out before frame content completely builds up[cite: 1].
  • Flexible Scene Splitting and Merging: If the Auto-Split automation fails to cleanly track an interface change point, you can manually prompt a target into becoming an independent "new scene" from the lower thumbnail strip[cite: 1]. Conversely, triggering the [Merge] option button immediately maps all subsequent macro actions cleanly onto a singular shared image trigger node[cite: 1].
  • Final Test and Saving: Click the [Test Run] button to execute and confirm that your entire logic macro successfully runs without issues[cite: 1]. Selecting [Save & Close] wipes out any stray, unassigned peripheral frame captures left in the thumbnail cache folder, automatically keeping your macro file footprints optimized[cite: 1].
  • Execution in the Main UI: Adjust the operation mode selector to [Extended] to step away from default single clicks and instead fire your compound macro chain events[cite: 1]. You can instantly open your macro workspace and adjust configurations anytime by clicking the [Advanced] settings action trigger[cite: 1].

Release Notes: Aimeck15 v1.2 (V2 Core Update)

Welcome to the Aimeck15 v1.2 release! This update introduces the highly anticipated V2 Core Engine, bringing a completely overhauled UI for advanced users, state-of-the-art anti-detection features, and absolute stability for Linux (Proton/Steam) environments.

1. Intelligent UI/UX Refinement for Click & OCR Settings

We completely redesigned the layout structure of the detailed settings panel. By maximizing the connectivity between the image preview (top) and detailed settings (bottom), we achieved an intuitive, high-speed workflow for heavy users. This is not just a visual upgrade—it integrates automatic state confirmation and bi-directional real-time rendering.

v2アメバ用OCR設定画面

(Image: Aimeck15 Main Screen showcasing OCR setup)

  • Automatic Operation Mode Switching & Row Highlighting: The 3rd (Click), 4th (ROI), and 5th (OCR) setting rows are now fully wrapped in TouchArea components. Touching any row instantly sends a request_mode_change signal to the Rust backend, brightly highlighting the row (rgba(255, 255, 255, 0.1)). This synchronizes active settings with the preview area's mode (0=Click, 1=ROI, 3=OCR) seamlessly, eliminating the cognitive load of searching for manual mode toggle buttons.
  • OCR Button Consolidation:
    The previously isolated "OCR Preview" (btn_ocr_preview) and newly added "OCR Settings" (cfg_ocr_settings) buttons have been beautifully consolidated into the right side of the 5th row. Checking the enable box, editing target values, previewing, and opening detailed settings can now all be completed in a single horizontal line.
  • Invisible Range Commitment (Automatic Save): We eliminated raw "Apply/Commit" buttons from the UI. Clicking the "OCR Settings" or "OCR Preview" buttons now triggers the ocr_rect_btn_clicked() signal internally before launching dialogs. This automatically crops and saves the image data in the background, entirely solving the issue of users forgetting to apply selections.
  • Real-Time Range Synchronization (Bi-directional UX): Main screen click locks are now fully released even when the OcrSettingsDialog is active. Dragging or re-selecting a region on the main screen (on_ocr_rect_selected) directly hooks into the new ocr_ops::update_live_preview function.
    • The Flow: The Rust backend detects the drag, recalculates offsets against the CACHED_FULL_IMAGE, and instantly regenerates the cropped preview image. This updates the active dialog's preview_image in real time, simultaneously saving coordinates to app_config.json.
    • Result: You can leave the dialog open and dynamically fine-tune your selection bounds on the main screen.

2. Anti-Detection "Random Delay" Integration & Full Internationalization

We moved away from mechanical, fixed-interval clicking by introducing a core engine Humanizer that precisely simulates human operational variance. This robust update goes beyond basic wait loops, ensuring strict layout protection across all supported languages and integrating deeply with our dynamic i18n logging engine.

v2アメバ用クリック設定欄拡大画面

(Image: The new Click Settings panel featuring the Random Delay configuration)

  • Human-like Random Wait Logic: A random sleep generator is now embedded directly into the action flow of action_ops.rs (before both standard clicks and delay-re-recognized clicks). It generates a random millisecond delay between 0.00 and the user-defined Max seconds before executing humanized_move_and_click. This drastically improves evasion against aggressive pattern-matching bot detection systems.
  • Layout Protection & Text Expansion Mitigation: We adopted the concise term "Random Delay" for the new UI labels. To prevent text expansion clipping in European languages (German, French, etc.), we optimized the CSV translation data using abbreviations (e.g., "Rnd"), ensuring all labels fit within a strict 10–12 character limit across 11 languages.
  • i18n-Compliant Dynamic Value Logging: Terminal outputs now correctly display the exact randomized wait time (e.g., ⏳ [Random Delay] A.png: Waited 0.18s before clicking). This is achieved cleanly using %s placeholders conforming to Aimeck's i18n engine, rather than relying on crude .replace string hacks in Rust.
  • Fail-safe Logging Multilingualization: Hardcoded Japanese warning logs (e.g., "🔧 復旧設定ロード...") deeply embedded within the lifecycle_manager.rs stack detection and recovery flow have been extracted into the i18n CSV. Users worldwide will now receive clear, native-language explanations when the fail-safe is triggered.

3. Complete Freeze Recovery Reconstruction & "Zero-Overhead" Monitoring for Linux

We fundamentally resolved the process tree zombie issues specific to Proton/Wine environments during the target application freeze-detection and reboot cycle (Lifecycle Manager). Furthermore, the process monitoring algorithm has been overhauled to achieve ultimate stability without sacrificing Aimeck's signature ultra-lightweight, 80fps+ image recognition performance.

  • Proton/Wine Residual Lock Clearance: To combat Linux-specific hang-ups, we integrated a proprietary cleanup sequence (including wineserver -k) that runs immediately after sending SIGKILL to the target PID. This completely eliminates the fatal silent errors where Steam blocks double launches due to hanging emulation layers.
  • Zero-Overhead Native OS Detection (O(1) Check): We deprecated the heavy full-process scan via sysinfo. Instead, the system now directly queries the Linux filesystem (/proc/<PID>). This $O(1)$ complexity check drops monitoring overhead to near zero, silencing OUT_OF_BOUNDS warnings during boot timing conflicts and ensuring flawless background execution.
  • Transparent Fail-safe Behavior via Localized Logging: Internal recovery statuses and countdown displays are now freed from hardcoded strings and fully integrated into the i18n CSV framework (utilizing %s placeholders). Detailed logs such as "✅ Process termination confirmed" now render beautifully in the user's native language, providing absolute peace of mind during long unattended sessions.
Assets 4
Loading