**OPERATING MODE: 1 — Tree-First** *User provided [INPUT\_AVAILABLE]: tree. No file contents are visible. Lineage will be speculative until priority files are pasted or uploaded.* **GRANULARITY: Table level** *Default selected — no column-level content is visible to support finer granularity.* **TRACE DIRECTION: Both (upstream + downstream)** **FILES/DATASETS TRACED (Direct):** None **FILES IN TREE ONLY (Speculative):** None yet provided **FILES/DATASETS NOT AVAILABLE:** All — no project tree has been pasted into this conversation. --- ## ⚠️ FALLBACK MODE TRIGGERED The protocol requires at least the **project tree** to operate in Mode 1, but no tree output has been included in your message — only the variable declarations: ``` [TECH_STACK]: Python / dbt / Snowflake [LINEAGE_SCOPE]: raw.orders → marts.revenue_report [INPUT_AVAILABLE]: tree [TRACE_DIRECTION]: both ``` I will not fabricate a lineage chain, table names, transformation logic, or dbt model paths for `raw.orders → marts.revenue_report` without visible evidence. --- ## To Proceed, Please Provide One of the Following **Option A — Project Tree (Mode 1, recommended for large projects)** Paste the output of one of these commands from your dbt project root: ```bash # Option 1 — Standard tree tree -L 4 --dirsfirst # Option 2 — If tree is not installed find . -type f \( -name "*.sql" -o -name "*.yml" -o -name "*.yaml" -o -name "*.py" \) | sort ``` Once I see the tree, I will identify the 5–10 priority files most likely to sit on the `raw.orders → marts.revenue_report` lineage path and request those specifically. **Option B — Paste Key Files Directly (Mode 3, fastest)** Paste the contents of any combination of: - The dbt model that produces `raw.orders` (likely a source definition or a staging model) - The dbt model(s) that produce `marts.revenue_report` - Any intermediate models between them (e.g. `int_orders`, `fct_orders`) - Your `sources.yml` or `schema.yml` files **Option C — Upload Files (Mode 2)** Upload the relevant `.sql` and `.yml` files directly if your platform supports it. --- Once input is available, I will produce the full eight-phase lineage analysis: lineage graph, health metrics, trust levels, break point classification, and a risk-tiered remediation plan.