Project tracking recipes for AI agents
Issues, tasks, and project management
Your agent files the work item; the credential stays home. Every recipe on this shelf delivers into a tracker's API through a standing pipe: the agent posts an intent, a transformation shapes it to the tracker's grammar, and the delivery is signed server-side with a secret the model never sees. What comes back is a receipt with the created item's id, so the agent can quote what it did instead of claiming it.
The trackers differ more than their marketing admits. Azure DevOps takes batched creates and updates through one pipe, so it gets a family: file, query, read, manage. Jira wants a broader token scope than its error messages let on. Linear speaks GraphQL. Each recipe carries its provider's quirks so you do not rediscover them at install time.
Start with the recipe for your tracker below. One install is one pipe: it fits the smallest plans, and the calculator on each recipe page shows exactly what it costs yours.
Create and update Azure DevOps work items in one batched pipe: priority, state, tags, and sprint included.
Query Azure DevOps work items with WIQL and read the matching ids back, without your AI ever holding the PAT.
Read fields for a batch of Azure DevOps work items by id: title, state, tags, priority, iteration.
Let your AI file Azure DevOps work items (bugs, tasks) the model never holds credentials for.
Your team's AIs flag the git pushes their humans care about and stay silent otherwise. Filtering by judgment, not rules, on one shared labeled stream.
Let your AI file GitHub issues with a fine-grained PAT it never touches.
Search GitHub issues and read whole threads back, without your AI ever holding the PAT.
Update, close and reopen GitHub issues in batches, with the PAT held server-side.
Let your AI file Jira Cloud issues with an API token it never touches.
Let your AI file Linear issues with an API key it never touches.
Let your AI add Todoist tasks; the API token never enters the conversation.