Kagan - AI Orchestration Layer

Get a task by ID.

The response includes a board_hint field summarizing other active tasks in the same project so the agent can decide whether to call task_list() for coordination.

Create multiple tasks at once.

Each entry must have a title key and may include description, priority, base_branch, acceptance_criteria, and agent_backend. Returns the list of created tasks.

Check whether a task has an interactive session.

Start an interactive session for a task.

Get the latest task and session status for a task.

Cancel a task run by task id.

Detach from an interactive session and update task state.

Approve a review-ready task.

Reject a review-ready task with feedback.

Merge an approved task into its base branch.

Rebase a task branch onto its current base branch.

Report the AI review verdict for a single acceptance criterion.

Call this once per criterion during review, BEFORE calling review_approve or review_reject. verdict must be 'PASS' or 'FAIL'. reason is a one-line justification.

Clear all AI review verdicts for a task. Call before starting a new review.

Sync external items from a plugin source into the active project.

Imports issues from the specified repository as kagan tasks. Labels like priority:high and kagan:auto on GitHub issues auto-map to task properties. Operation is idempotent — previously synced issues are skipped.

Args: plugin: Plugin to sync (e.g. "github"). Use plugins_preflight to list available. repo: Repository in owner/repo format (e.g. "octocat/hello-world"). state: Issue state filter — "open", "closed", or "all". import_label: Only sync issues with this label.

Check if a plugin's external dependencies are satisfied.

Returns pass/warn/fail checks for the requested plugin (or all plugins). For github: verifies gh CLI is installed and authenticated.

Args: plugin: Plugin to check. If omitted, checks all installed plugins.