ticket-swarm-workflow

Use when live product annotations or reviewer notes must become hard reconciliation rows before a goal loop, ticket swarm, or QA pass may claim readiness.

Use when a merged Squad slice or Squad-style release candidate is ready for delivery work: deploy-readiness verification, artifact and manifest checks, updater feed verification, release-note assembly, human-triggered release handoff, and rollback or re-release receipts.

Use when today's real calendar availability needs to become a concrete ticket execution calendar backed by the issue-tracker dependency graph.

Use when today's execution day is already in motion and calendar blocks, issue-tracker due-work, and review waves need to be reflowed because a blocker cleared, a lane slipped, or review landed differently than planned.

Use when turning a live browser-backed design review into durable Design Contract tickets, then handing approved route clusters into implementation-contract compilation.

Use when feature ideas are scattered across team chat, meetings, and an issue tracker and we need one durable ledger for beta thesis, explicit non-goals, parking-lot features, kill-list candidates, and undecided gaps.

Use when a product owner gives a /goal-style long-running product outcome that should loop through discovery, implementation-contract materialization, readiness QA, telemetry review, and blocker re-planning.

Use when approved Squad Design Contract clusters or implementation-readiness rows need to become seam-locked child contracts under the active Implementation Contract parent before ticket-to-human-review execution.

Use when a daily execution plan needs to become real issue-tracker ownership, due-date, and state-allocation truth for dependency-ready child issues.

Use when a user enters a mission-shaped flow and Squad must make legible context visible before mission definition by building or refreshing an ontology/epistemology context graph from code, connectors, transcripts, memory, and elicitation.

Use when Squad has enough discovered context to turn a user outcome into a canonical mission definition, task DAG, route/control posture, governance authority packet, and Approve & Run handoff.

Use when rendering, auditing, or summarizing the mission overview surface or generated mission card from canonical mission operating context.

Use for an active mission task when Squad must execute, review, verify, and receipt backend, frontend, or devops work autonomously against mission requirements.

Use when multiple dependency-ready tickets can advance simultaneously and we need explicit execution waves, lane ownership, and serialized-resource rules before work starts.

Use when multiple tickets are expected to converge into one shared human-review wave, followed by patch and re-review loops that should stay synchronized instead of ad hoc.

Use when a bounded ticket set is already implemented and verified enough to enter one-by-one human signoff. Selects the review order, opens each surviving worktree in sequence, runs the human QA loop, captures verdict, merges verified work into main, reruns mainline proof, updates `issue_tracker`, and only then advances to the next ticket.

Use when Squad agent behavior, prompts, tool manifests, eval traces, trajectories, datasets, or applied-AI harness changes need to be engineered through source traces, golden/regression/challenge/canary cases, scorers or judges, and regression receipts before product claims or implementation closeout.

Use when a Squad execution day has real board pressure, calendar constraints, or schedule-slip correction and we need a calendar-aware close plan before ticket execution starts.

Use when a Squad ticket is already through automated review and verification and needs a human product-QA runtime. Prepares the surviving worktree, launches the right dev surface, provides the human review checklist, and routes any findings back into the owner ticket loop without using root main as the review workspace.

Use first when a Ticket Swarm goal, ticket, annotation set, or execution day needs the smallest correct next workflow skill selected without loading the whole plugin.

Use when an implementation contract needs human signoff before development/testing begins. This verifies the contract, not the implementation.

Use after the configured product reviewer has annotated a UI failure pattern and future agents should proactively find, classify, and route the same class of product-surface abstraction mistake without requiring repeated manual annotation.

Use when a ticket will execute with parallel mutating lanes and needs explicit git worktree and branch isolation before multi-agent execution starts.

Use after lane review and verification to collapse parallel lane work back onto one surviving integration branch/worktree and record cleanup posture before human verification and the later merge-to-main close gate.

Use when a completed or terminated mission enters human review and Squad must verify mission and task outcomes, surface evidence, collect feedback, and preserve learning without treating human attention as missing machine proof.