atsscorelisted

# ATSScore (Pipeline Phase 2b) Score resume fit for each `to apply` job using **sheet data only** (title + Skills in Details) and local PDF resumes. **Writes:** | Col | Header | Content | | --- | --- | --- | | **I** | **ATS score** | Numeric scores per resume label | | **J** | **Suggestions on Resume** | Rule-based improvement tips for best-matching resume | No job URL fetch. No LLM. No browser. Runs **in parallel with LeadGru** (Phase 2) — LeadGru uses browser for G/H; ATSScore uses one Python CLI for I/J. ## Pipeline mode (default) **When:** Jobgru Phase 1 verified sheet append. Triggered automatically alongside LeadGru — no separate user prompt. **On/off:** | Condition | Action | | --- | --- | | No PDFs in `data/resumes/` | Skip; set `atsscore_status: skipped` in run JSON | | User prompt contains `ATS scoring: no` | Skip | | Resumes present + not disabled | Run | **Scope:** All rows where Status = `to apply` AND column **I (ATS score)** is empty — includes new rows and backfill of older rows. ```bash .venv/bin/python scripts/ats_score.py score --all ``` Optional dry-run first: ```bash .venv/bin/python scripts/ats_score.py score --all --dry-run ``` ## Standalone mode (backfill only) **When:** User explicitly asks to run ATS scoring without Jobgru/LeadGru. Use [prompts/atsscore-run.md](../../prompts/atsscore-run.md). ```bash .venv/bin/python scripts/ats_score.py score --all ``` ## Prerequisites Run **Jobgru check**: `.venv/bin/python scripts/jobgru_check.p