ai systems / Engineering project
Atelier360 — Agent Harness & Workshop Operations
A desktop workshop application with a safety-conscious agent harness
Atelier360 is a Rust and Tauri desktop application for workshop operations, combining commercial and CAD workflows with an AI assistant. Its agent harness coordinates tool calls, user confirmation, recovery, and bounded context while preserving application authority over business actions.
Project status: In development; release verification ongoing (candidate-reported).
Architecture description provided by Mourad. Ask for a demo, code walkthrough, or supporting artifacts.
Architecture & decisions
- Approval boundaries: consequential business actions require user confirmation; atomic claims and durable refusals help prevent duplicate execution.
- Recovery and reconciliation: for covered operations, durable receipts identify committed actions after interruption; unknown outcomes are not replayed automatically.
- Context and privacy controls: scoped result storage, bounded reads, and outbound-data policies for remote model requests.
- Tool execution: a resource-bounded Code Mode runtime and deterministic checks support structured workflows.
Technology
What to explore in a walkthrough
- Where the system runs today and which parts are production, development, or research.
- The architectural tradeoffs, failure modes, and tests behind these decisions.
- What AI agents contributed and how their output was reviewed.
- Operational evidence and measurable results available for review.
Project descriptions and experience are candidate-provided. This résumé's MCP tools are directly testable; other systems are available for an architecture walkthrough. Current and past employment positions and employers are private. Professional focus and target roles do not disclose a current job. Independent verification and references require direct contact.