Developer releases Current, an AI agent change control tool
Understand this story
Developer santosh_ycomb has released the open-source tool Current (implemented in Elixir), which restricts AI agent output to nine types of typed intent proposals; proposals must be approved by a human or another agent before the runtime separately executes apply, and the entire process is written to a hash-chain log. The tool is currently at an early stage and supports only single-node and local projects. The author notes that Current is not a sandbox or an authentication system: approved commands execute with the runtime's own permissions; the runtime can start without an API key set, but the model catalog will be empty; MCP does not verify caller identity, so anyone who can access the endpoint can submit, approve, and apply proposals.
AI overview based on source reports · Updated 6 min ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Hacker News AIShow HN: Current – Change control for AI agents (Elixir)
Current is an Elixir runtime in which AI agents cannot act directly: they submit typed proposals that are validated, reviewed, approved, and only then applied by a separate trusted interpreter. Every step, from the request and the files the model read to approvals and effect outputs, is recorded in a hash-chained, event-sourced log that can be replayed to rebuild state. It is early-stage, single-node, and explicitly not a sandbox or authentication system.
Activity for this story
Not enough continuous observations to show a trend yet.