Axleta Context-Packer

Official Interactive User Manual & Reference Guide

← Back to App

1. Product Overview

Axleta Context-Packer is a developer productivity utility engineered to serialize, aggregate, and format entire GitHub codebases into token-optimized LLM context prompts. It parses full directory trees and source files into clean XML structure maps compatible with Anthropic Claude 3.5, OpenAI GPT-4, and Google Gemini models.

Application Interface Overview
Header Interface Navigation
Figure 1.1: Axleta Context-Packer Navigation Bar featuring account state, credits counter, and dark/light mode toggle.

2. Authentication & GitHub Integration

Accessing private repositories or managing packing credits requires account authentication. Context-Packer supports standard email/password authentication alongside single-click GitHub OAuth integration.

Login Window Option
Figure 2.1: User Login authentication modal.
Sign Up Window Option
Figure 2.2: New Account Creation modal with automatic free context packs.

3. Repository Configuration & Settings

Configure source repository targets, personal access tokens for private repositories, file exclusions, and context extraction parameters within the Repository Configuration panel.

Repository Configuration Panel
Figure 3.1: Repository Configuration options and extraction controls.

Key Configuration Fields:

  • GitHub Repository URL: Paste the target URL (e.g., https://github.com/expressjs/express).
  • Personal Access Token (PAT): Required for private repositories. Remains stored strictly client-side inside browser memory.
  • Pack Repository Context: Triggers recursive API fetching, dynamic file filtering, and token extraction.

4. Globs Exclusion & AST Skeleton Parsing

Optimize context windows and eliminate noise by applying Dynamic .gitignore filtering and custom exclusion patterns.

1. Dynamic Glob Exclusions

Comma-separated patterns to omit compiled binaries, dependencies, or tests. Example default patterns: *.min.js, docs/**, **/*.test.ts.

2. AST Structural Parsing (Skeleton Extract Mode)

When enabled, source code is parsed into Abstract Syntax Tree (AST) skeletons. Implementation details are stripped while retaining exported class definitions, interface signatures, and method headers—reducing overall token consumption by up to 70%.

**If a developer needs an AI model to fix a bug or write an integration, the AI must see the full code implementation inside lib/ to do its job then leave these two options and you don't need to select that 1, 2. It packs 100% of the core app logic (lib/, controllers, routes) so the AI can read, debug, and write functional code.**

5. Packed Prompt Output & Model Compatibility

The generated output is automatically calculated for token length and formatted using XML tags to ensure zero ambiguity during LLM multi-file reasoning.

Packed Prompt Output Window
Figure 5.1: Packed prompt window displaying real-time token count, LLM fit status badge, and XML output canvas.

6. Credits System & PayPal Recharges

Every repository packing operation consumes 1 credit from your user account. New users receive 3 starter credits upon signup. Additional credits can be purchased via instant PayPal checkout.

Credit Recharge Modal
Figure 6.1: PayPal credit top-up options for Starter and Developer packs.

7. Support & Further Assistance

Need help integrating with custom CI/CD pipelines, encountering rate limits on large monorepos, or requiring custom enterprise tiers?

Axleta Developer Support
Contact our engineering team for assistance or bug reports.
Email Support