global-impact-tracker 0.1.0

Local-first CLI for tracking AI-assisted work impact and generating STAR stories.

Global Impact Tracker

Global Impact Tracker is a local-first CLI for proving the value of AI-assisted work. Log tasks, capture metrics, and generate interview-ready STAR stories — all from the command line, with no hosted backend required.

---

Free CLI

Install

pip install global-impact-tracker

Log a task

impact-tracker log \
  --project "Payments API" \
  --task "Refactored auth middleware to remove legacy session token storage" \
  --baseline-hrs 4.0 \
  --ai-sec 420 \
  --status "Success" \
  --task-type refactor \
  --complexity high \
  --tools-used "claude|windsurf"

Logged run to: /Users/you/.impact_tracker/global_productivity.csv

Capture metrics

impact-tracker metrics

{
  "queries_processed": 12,
  "total_hours_saved": 18.4,
  "latency_reduction_pct": 91.2,
  "success_rate_pct": 100.0,
  "projects_count": 3,
  "system_health": "healthy"
}

Generate the dashboard

impact-dashboard

Opens live_impact_dashboard.html in the current directory — a visual breakdown of projected vs. actual hours by project.

---

MCP Server (free for a limited time)

The paid MCP server brings AI-powered STAR story generation, decisions capture, and richer reflection directly into your AI coding environment.

It is currently free. Request access and a license key will be emailed to you.

Request access

Request a free MCP key →

You will receive a license key (gip-...) and full installation instructions by email, typically within 24 hours.

Install the MCP server

pip install global-impact-tracker-mcp

You will also need a Gemini API key (free tier is sufficient):

export GEMINI_API_KEY="your-gemini-key"
export IMPACT_TRACKER_LICENSE_KEY="gip-your-key-here"

Configure your client

Pick the tab for your AI coding environment.

Claude Code

Add to ~/.claude/claude_desktop_config.json:

{
  "mcpServers": {
    "impact-tracker": {
      "command": "impact-tracker-mcp",
      "env": {
        "IMPACT_TRACKER_LICENSE_KEY": "gip-your-key-here",
        "GEMINI_API_KEY": "your-gemini-key"
      }
    }
  }
}

Restart Claude Code, then confirm the server is running:

/mcp

Codex CLI

Add to ~/.codex/config.json:

{
  "mcpServers": {
    "impact-tracker": {
      "command": "impact-tracker-mcp",
      "env": {
        "IMPACT_TRACKER_LICENSE_KEY": "gip-your-key-here",
        "GEMINI_API_KEY": "your-gemini-key"
      }
    }
  }
}

Gemini CLI

Add to ~/.gemini/settings.json:

{
  "mcpServers": {
    "impact-tracker": {
      "command": "impact-tracker-mcp",
      "env": {
        "IMPACT_TRACKER_LICENSE_KEY": "gip-your-key-here",
        "GEMINI_API_KEY": "your-gemini-key"
      }
    }
  }
}

Windsurf

Open Settings → MCP Servers → Add Server and enter:

Field	Value
Name	impact-tracker
Command	impact-tracker-mcp
Env: IMPACT_TRACKER_LICENSE_KEY	gip-your-key-here
Env: GEMINI_API_KEY	your-gemini-key

Sample prompts

Once the MCP server is running, paste any of these into your AI client to get started:

Log a task: I refactored the auth middleware today, it would have taken me about 3 hours manually but only took 8 minutes with AI

Show me my impact dashboard

Generate a STAR story for my work on the payments project — I'm preparing for a senior engineer interview

Generate a cross-project STAR story covering all my recent AI-assisted work

What engineering decisions did I make this week?

How many hours have I saved this month across all projects?

---

How it works

1. Log tasks as you complete them — baseline hours (what it would have taken manually) and actual AI-assisted seconds.
2. Metrics are computed locally from your CSV log at ~/.impact_tracker/global_productivity.csv. No data leaves your machine on the free tier.
3. STAR stories (paid) are generated by Gemini using your real metrics and captured engineering decisions — grounded in actual numbers, never fabricated.

---

Public package boundary

The public package contains:

• shared tracker logic and CSV-backed storage
• metrics generation and the dashboard HTML generator
• the placeholder entitlement verifier interface
• the impact-tracker and impact-dashboard CLI entry points

The public package does not ship the MCP server, internal key issuance tooling, or any real signing secret. Those live in the private global-impact-tracker-mcp companion repo.

---

Environment variables

Variable	Where	Purpose
IMPACT_TRACKER_LICENSE_KEY	Customer	Activates Pro MCP features
GEMINI_API_KEY	Customer	Powers STAR story generation and decisions capture

---

Built by Mass Ave Labs