Output Template (Locked)
Deterministic output shape. Every delivery renders these sections in this order. Never skip a section; render "None applicable" if a section has nothing to report.
Template
## Summary
- **Data source audited:** {datasource_name}
- **Geographic fields found:** {N}
- **Findings by severity:** Critical {N_c}, Warning {N_w}, Note {N_n}, Info {N_i}
- **Sample-size status:** {all_verified | mixed | all_unverified}
## Findings by Severity
### Critical ({N_c})
For each Critical finding:
- **{G_id}: {field_name}** — {current_role} → recommended {recommended_role}
- Evidence: {percent_or_sample_values}
- Sample size: {N_rows} ({verified | unverified})
- Fix: {Desktop_UI_steps}
- {optional_precedence_note: "Secondary to G{n} on same field"}
### Warning ({N_w})
Same format as Critical.
### Note ({N_n})
Same format.
### Info ({N_i})
Same format.
## Handoff (if applicable)
- **Build a map next?** Trigger `worksheet-builder` with: "build a map from {geographic_field_name}."
- **Performance tuning next?** Trigger `extract-optimizer` with: "optimize the extract for {datasource_name}."
- **Pre-ship hygiene next?** Trigger `publish-readiness` with: "publish readiness check for {workbook_name}."
- **Not sure what's next?** No handoff. Skill's job ends at findings delivery.
## Handoff YAML (optional, schema_version geo_audit_handoff/1)
Only render if a downstream skill has requested a machine-parseable handoff.
```yaml
schema_version: geo_audit_handoff/1
datasource_name: {name}
findings:
- g_id: {G1 | G2 | ... | G6}
field: {field_name}
current_role: {role_name}
recommended_role: {role_name}
severity: {critical | warning | note | info}
evidence: {string}
sample_size: {N}
verified: {true | false}
summary_counts:
critical: {N_c}
warning: {N_w}
note: {N_n}
info: {N_i}
verified: {true | false}
notes: {optional_free_text}
## Filling Rules
- Every `{placeholder}` must be filled or replaced with "not applicable" / "none." Never leave a placeholder.
- Every severity section header renders even with zero findings ("Critical (0)" with "None." body). Silence would be ambiguous.
- Sample-size status in Summary is `all_verified` (all findings ≥1000 rows), `mixed` (some verified, some not), or `all_unverified`.
- Handoff section renders three routing options plus the "no handoff" fallback.
- Handoff YAML renders only when downstream skill requested.
## What NOT to Render
- Tool names (`get-summary-data`, `list-available-fields`). Stay user-facing.
- Internal reference-file names (`geocoding_checks.md`, `severity_rubric.md`).
- Grader IDs, validation-gate IDs, or hard-constraint IDs.
- Hedged language ("might be misassigned," "could be ambiguous"). Findings cite G-ID with evidence; severity is locked.
## Refuse to Deliver on Validation Gate Failure
If Phase 4 fails V1, V2, or V3, do NOT render this template as-is:
- **V1 fail (finding without G-ID):** Drop the finding. Never render an unlabeled geographic finding.
- **V2 fail (severity off-rubric):** Recompute severity from rubric. If ambiguous, pick higher severity.
- **V3 fail (no sample size):** Flag the finding as "unverified" rather than dropping.
The Validation Gate is a hard block, not a soft warning.