executing-plans
executing-plans
Use when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available
workflows-queuesWork with InvokeAI workflow library records, default workflow JSON assets, session queue items, batch payloads, queue privacy redaction, and workflow/queue API behavior.
Independent GitHub snapshot · Automated classification · Original documentation governs
Use this sub-skill when the task involves InvokeAI workflow JSON/library records, default workflow assets, queue items, batch enqueue payloads, workflow thumbnails, public/private workflow behavior, or multiuser queue redaction.
references/workflow-records.md for workflow DTOs, library storage, CRUD, thumbnails, tags, categories, opened_at, and public/private access rules.references/session-queue-and-batches.md for Batch, BatchDatum, NodeFieldValue, SessionQueueItem, status lifecycle, GraphExecutionState serialization, retry/cancel/clear behavior, and redaction.references/default-workflows.md for bundled default workflow JSON shape, startup sync rules, and safe default-workflow inspection.references/troubleshooting.md for common workflow/queue failures and exact diagnosis paths.workflow-nodes for custom node implementation, invocation schemas, graph construction, and node field semantics beyond queue payload validation.model-management for model identifiers, model nodes, and compatibility warnings in default workflows.operations-config for auth setup, multiuser configuration, admin roles, and deployment-level queue settings.scripts/inspect_default_workflow.py <workflow.json> checks default workflow JSON shape, node/edge/exposed-field references, category, tags, and summary metadata without importing InvokeAI.scripts/validate_batch_payload.py <payload.json> checks queue enqueue batch structure, zipped batch lengths, homogeneous item types, duplicate node/field mappings, node presence, and obvious field mismatches.Keep workflow JSON and queue records conceptually separate: workflow files are frontend/library records, while queued session data is serialized execution state derived from an invocation graph.