TrustMeImWorking

What Problem It Addresses

Some AI usage reviews treat token-call trends as evidence of active work. When quota is allocated daily or weekly but real tasks arrive irregularly, the account can look underutilized or bursty. TrustMeImWorking is a straightforward tool for that gap: it spreads an approved token budget across scheduled windows so the call pattern resembles ongoing work rather than a one-off spike.

How It Works

It runs as a background daemon and calls a configured LLM API at controlled intervals:

• Work Simulation: fires during weekday work hours with job-related prompts such as code review, architecture design, and data analysis.
• Spread: distributes the day's budget evenly across the remaining hours for a smoother consumption curve.
• Immediate: starts consumption right away and resets at 00:00 the next day.

Setup is usually generated by an interactive wizard. It supports OpenAI, DeepSeek, Anthropic, Gemini, Groq, and OpenAI-compatible local or enterprise endpoints. Key fields include Base URL, model, API key stored in local config.json, weekly token budget, job description, and work hours. After startup, a dashboard refreshes every two seconds, and operators can inspect logs, status, run a single session, or use dry-run simulation.

Limits and Caveats

It addresses usage pacing, not business output. If an organization requires actual tasks, code, or documentation, this tool does not replace those deliverables. API keys should be treated as local secrets, and enterprise gateways may need proxy, JWT, or mTLS settings.