PitchHut logo
Streamlined collaboration for human-centric development teams.
Pitch

The IPIVC standard offers a structured approach to collaboration among developers, architects, and QA teams working on the same codebase. Integrating AI assistance, it ensures real-time conflict detection and promotes efficiency through a systematic protocol that governs interaction and task management.

Description

IPIVC is a collaborative standard designed for Human-Centric Teams comprising Developers, Architects, and QA professionals working on a shared codebase with AI assistance. This repository provides a well-structured framework enabling teams to manage and coordinate their development efforts effectively, ensuring better quality and reduced conflict.

Key Features

  • Hybrid Architecture v2.0: IPIVC integrates two layers of security:
    1. Git (Persistence): Enables file locking visible to the whole team.
    2. LokiVector (Semantics): Acts as a "Central Oracle," detecting conflicting intentions (e.g., "Fix Auth" vs "Login Refactor") in real-time.

Daily Workflow Configuration

Teams can enhance their workflow through a two-window setup:

  1. Window 1 (IDE + AI): Where coding takes place with an AI Agent (e.g., Trae, Cursor).
  2. Window 2 (Dedicated Terminal): Where coordination tools are executed.

The IPIVC Protocol as a "Human Contract"

The protocol follows several critical steps to facilitate team collaboration:

  1. Investigate (Sync): Run python tools/session_manager.py to check who is working on what within the team.
  2. Plan (Lock & Spec): Reserve the workspace and define Functional Specifications in the session file.
    • Example: "The login must lock after 3 failed attempts." This locks resources for the team.
  3. Implement (AI Execution): Generate context using python tools/context_generator.py and feed it to your AI, which understands which files can be accessed.
  4. Verify (Auto-Audit): Execute python tools/verify_task.py to check whether:
    • Scope: Did the AI access prohibited files?
    • Specs: Are functional requirements met?
    • Tests: Do automated tests pass? A feedback report (AI_FEEDBACK_REPORT.md) will be generated for further refinement.
  5. Correct (Loop): Deliver the report to the AI, iterating until all criteria are satisfied, after which the lock is released, and changes are pushed.

Project Structure

  • server/: Hosts the LokiVector server for semantic conflict detection.
  • tools/: Contains utility scripts like:
    • session_manager.py: Coordination client.
    • verify_task.py: Quality and scope auditor.
    • context_generator.py: Bridge for AI.
  • memory/: Knowledge database of the project.
  • AGENTS.md: Open standard for Agent instructions.

For complete documentation, visit the Full Documentation.

IPIVC aims to enhance collaboration among developers, providing clear guidelines and tools for effective teamwork.

0 comments

No comments yet.

Sign in to be the first to comment.