Dingdawg Planning Agent

A local planning assistant for sprint estimation, task breakdown and risk analysis that learns your team's velocity.

What you can do

The registry description presents it as a planning AI covering sprint estimation, task breakdown and risk analysis, and says it learns your velocity. The README describes a broader governed-agent platform: every action a governed agent takes generates a cryptographically signed receipt recording what it decided, why, which model, which version and at what time. Listed live capabilities include developer signup that issues a scoped API key, governed execution that routes agent calls through metering and audit, compliance classification, regulation lookup, agent trust score with a live stream, and Stripe billing. The manifest does not list the server's individual tools.

When to use it

Consider it when you want an assistant to help estimate sprints, break work into tasks and analyze delivery risk while adapting to your team's velocity. It may also suit teams that want agent actions recorded as signed receipts for audit or AI-regulation evidence. It is not aimed at general-purpose coding or data tasks.

Requirements

Runs as a local stdio process installed from the npm package dingdawg-planning-agent, so Node.js and npm/npx are needed. Desktop only; no web executable. Requires the secret environment variable DINGDAWG_API_KEY, described as an API key for paid tier access, obtainable free per the manifest. Network access to the provider's API is implied by the README's architecture and endpoints. No authentication method is declared in the manifest.