🎯

AIMED: a glass-box view of your AI coding context

AIMED (AI Monitoring & Execution Dashboard) is a local MCP server + dashboard that lets you inspect and curate the project context your assistant runs on.

Read documentation View repository

Three Core Benefits

AIMED makes the assistant’s working context visible and editable, so you can keep project decisions, progress, and patterns aligned over time.

💰

Cost Savings & Efficiency

  • ✓ Keep LLMs focused to prevent context drift
  • ✓ Reduce token usage with efficient context retrieval
  • ✓ Faster development with organized knowledge
  • ✓ Clean active context removes task baggage
👁️

Transparency & Control

  • ✓ Spot AI missteps and wrong assumptions
  • ✓ See inside the AI's decision-making process
  • ✓ Track project evolution over time
  • ✓ Visual representation of AI context usage
🎯

Better Outcomes

  • ✓ Consistent AI behavior across sessions
  • ✓ Knowledge retention between sessions
  • ✓ Team collaboration with shared context
  • ✓ More reliable AI responses

Token efficiency (and fewer re-prompts)

AIMED helps reduce wasted tokens by keeping the assistant’s working context accurate, current, and easy to correct when it drifts.

Where the savings typically come from

  • ✓ Fewer correction loops: less “try again” prompting when the AI is missing a decision or repeating a wrong assumption.
  • ✓ Cleaner active context: remove stale task baggage so the assistant stays focused on what matters now.
  • ✓ Explicit project knowledge: decisions, progress, patterns, and custom data remain accessible and linked over time.

Practical workflow

  1. AI assistant produces an answer that’s off-target (missing context, wrong assumption, stale plan).
  2. You inspect the relevant context/entries in AIMED (graph, decisions, progress, contexts).
  3. You correct or update the “project truth”.
  4. Next request runs with the corrected context.

Core features

AIMED combines a ConPort MCP backend with a visual dashboard for inspecting and managing project context.

📊

Real-time Dashboard

Visual activity summaries and progress tracking with live updates.

  • • Recent activity feed
  • • Progress kanban boards
  • • Context change tracking
  • • Live polling updates
🕸️

Interactive Knowledge Graph

Visual exploration and creation of relationships between project elements.

  • • Drag-and-drop node creation
  • • Visual relationship mapping
  • • Interactive canvas
  • • Context menus for actions
🕒

Context Management

In-place editing for product and active contexts with version history.

  • • Visual form editors
  • • Version history tracking
  • • Change summaries
  • • Rollback capabilities
⚡

Decision Management

Visual CRUD operations for architectural decisions with tagging.

  • • Decision logging forms
  • • Tag-based organization
  • • Search and filtering
  • • Rationale tracking
✅

Progress Tracking

Kanban-style progress management with visual status indicators.

  • • Drag-and-drop kanban
  • • Task hierarchies
  • • Status automation
  • • Progress analytics
🔍

Advanced Search

Full-text and semantic search with visual result exploration.

  • • Full-text search (FTS5)
  • • Semantic vector search
  • • Visual result previews
  • • Cross-reference linking

Next steps

Review the code, read the documentation, and use the pages below as a starting point for evaluation, contribution, or collaboration discussions.

GitHub repository Documentation