Orchestrates structured external LCA packages (BAFU, USLCI) into TIDAS candidate rows via CLI conversion, queue state management, child skills, and deterministic publish handoff.
How this skill is triggered — by the user, by Claude, or both
Slash command
/lca-foundry-workflows:external-dataset-curated-importThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Use this top-level skill when the input is a packaged or semi-structured LCA dataset that should enter TianGong through a repeatable import profile.
Use this top-level skill when the input is a packaged or semi-structured LCA dataset that should enter TianGong through a repeatable import profile.
tiangong-lca dataset import-lca convert or tidas-tools..foundry/workspaces/<task-id>/.generic, bafu, or uslci.tiangong-lca dataset context-pack \
--type <process|flow|source|contact|unitgroup|flowproperty|lifecyclemodel> \
--profile ai-import \
--out-dir .foundry/workspaces/<task-id>/context/<type> \
--json
tiangong-lca dataset import-lca convert \
--input /abs/path/source-package \
--output-dir .foundry/workspaces/<task-id>/conversion \
--from-format auto \
--target tidas \
--json
Normalize converted rows into explicit support, flow, and process JSONL files.
Run schema validation and deterministic QA through the CLI.
Build the entity queue:
tiangong-lca dataset curation-queue build \
--processes .foundry/workspaces/<task-id>/rows/processes.jsonl \
--flows .foundry/workspaces/<task-id>/rows/flows.jsonl \
--support .foundry/workspaces/<task-id>/rows/sources.jsonl \
--out-dir .foundry/workspaces/<task-id>/curation-queue \
--json
tiangong-lca dataset curation-queue next \
--queue-dir .foundry/workspaces/<task-id>/curation-queue \
--json
Execute only the returned task. Support tasks complete before flow tasks; flow tasks complete before dependent process tasks.
flow-governance-review or a future flow-authoring alias.process-automated-builder or a future process-authoring alias.lca-publish-executor only after deterministic prewrite gates are ready.After each task finishes, write or update that task's checkpoint.json, then call curation-queue next again.
Before remote write, verify queue completion for the intended scope:
tiangong-lca dataset curation-queue verify \
--queue-dir .foundry/workspaces/<task-id>/curation-queue \
--type process \
--json
The verify report must be passed for the committed scope. If verify is blocked while next still returns ready, continue the queue.
Stop and return blocker evidence when:
tiangong-lca dataset curation-queue verify is passed for the committed scope.npx claudepluginhub tiangong-lca/skills --plugin lca-hybrid-searchGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.