What problem does it solve? When an AI-agent issue tree stalls, loops infinitely, or recovers too deeply, teams need a repeatable forensic procedure to find the exact stop point and define a general product rule instead of shipping one-off patches. ## Core Features & Use Cases - Tree Forensics: Walk the named issue tree node-by-node to identify the exact issue and state combination that stopped progress, with run ids and comment evidence. - Invariant-Checked Rule Design: Frame every fix as a general product contract that preserves three invariants: productive work continues, only real blockers stop work, and no infinite loops. - Approval-Gated Planning: Write a phased plan into the issue's plan document, open a request_confirmation interaction, and only decompose into child issues after board acceptance. - Use Case: A manager links a stalled issue tree and asks why work stopped; the skill produces a root-cause write-up, a contract-style product rule reconciled against doc/execution-semantics.md, and a phased remediation plan. ## Quick Start Ask the agent to diagnose why the linked issue tree stopped working and produce a root-cause analysis with a product-rule plan before any code changes.
