DEKC second brain
The knowledge bundle is a schema-typed second brain for data platform design work.
Schemas
python3 "${CLAUDE_PLUGIN_ROOT}/scripts/dekc_schemas.py" list
python3 "${CLAUDE_PLUGIN_ROOT}/scripts/dekc_schemas.py" intents
python3 "${CLAUDE_PLUGIN_ROOT}/scripts/dekc_schemas.py" validate --repo . --bundle knowledge
Intent query
# Design a report
python3 "${CLAUDE_PLUGIN_ROOT}/scripts/dekc_brain.py" "executive revenue" \
--intent design-report --repo . --bundle knowledge
# Land new data
python3 "${CLAUDE_PLUGIN_ROOT}/scripts/dekc_brain.py" "orders event hub" \
--intent land-data --repo . --bundle knowledge
# Metric design
python3 "${CLAUDE_PLUGIN_ROOT}/scripts/dekc_brain.py" "GMV" \
--intent design-metric --repo . --bundle knowledge
# Impact of changing a table
python3 "${CLAUDE_PLUGIN_ROOT}/scripts/dekc_brain.py" "gold order_daily" \
--intent impact --repo . --bundle knowledge --write
Workflow
- Retrieve with
dekc_brain.py(intent + query) → checklist + ranked concepts + 2-hop pack. For parent Q&A, spawn data-retriever (/dekc-retrieve) and keep the card only — do not dump the brain pack inline. - Design using only evidence from the pack (do not invent lineage)
- Capture new assets with
dekc_capture.pymatchingschemas/okf-concepts/ - Index is optional — search/pack refresh
knowledge/.dekc/index.sqlitethemselves.dekc_index.py buildisrefresh --force. - Optional grade if reverse-engineering
See schemas/README.md and docs/user_guide/user-guide.md.