de:wasserversorger-enrich

Journal entry appended. The run is complete.

Enriched 45 water suppliers this pass (23+22 shards), all research-only with no contact made and no captchas touched: 38/45 got full 5-core contact info, 33/45 emergency numbers, 39/45 an analysis URL or fallback, and 280 new municipality links were written. enrichment.jsonl grew 1484→1529, population coverage held steady at 99.99% (24 residual municipalities, 11,023 people). Several roster contradictions/discoveries surfaced for human review (a likely duplicate at Hille/WBV Hartum, a membership contradiction at WZV Woringen, expanded supply areas at WBV Usingen and Stadtwerke Bitburg, a new TLS host trap at Stadtwerke Halver) — all logged in the journal for the next run.

Picked up cleanly where the 05:30 run left off (enrichment.jsonl 1484, roster.jsonl 3375, roster-links.jsonl 18247 — matched the prior entry's reported finish exactly).

Batch build: build_enrich_batch_excl.mjs 45 enrich-20260903-1030 — 1916 pending (excluding 6 standing duplicate-mint pairs), 0 pending_domain (queue still empty). Split 23/22, dispatched to 2 synchronous general-purpose Agent subagents (run_in_background: false, both launched in one message, each explicitly told not to nest further sub-agents — neither did). Each wrote its own enrichment.part-shard{A,B}-enrich-20260903-1030.jsonl / roster-links.part-shard{A,B}-… shards.

Verified before merge: 0 malformed JSON across all 4 shard files, both shards' company_id sets matched their dispatched batch files exactly, 0 overlap between shard A and B, 0 overlap with the existing 1484-line enrichment.jsonl.

• Shard A (23): 21/23 full 5-core contact, 17/23 emergency_phone_number, 18/23 analysis_url or recorded fallback, 11/23 multi_zone, 114 fetches / 19 searches, 194 municipality links.
• Shard B (22): 17/22 full 5-core contact (5 misses are earned email gaps), 16/22 emergency_phone_number, 21/22 analysis_url or fallback, 8/22 multi_zone, 103 fetches / 13 searches, 86 new municipality links (6 companies already had complete links from prior passes, not re-added).
• Combined: 45/45 researched, no contact made, no captchas touched, 280 new municipality links.

Merge: enrichment.jsonl 1484 → 1529 lines / 1529 distinct company_id (+45 clean). roster-links.jsonl 18247 → 18527 (+280, exact match). map_roster.mjs ran clean (usual crop of no_ags_for_this_name unmatched Ortsteil/Verbandsgemeinde-composite names, nothing new). consolidate_roster.mjs: 177 shards / 3617 rows, still 3375 distinct companies; roster-links.csv 15749 distinct after dedup (+171 net vs the 05:30 run's 15578). coverage.mjs: population 99.99% → 99.99% (unchanged, expected for an enrich-only pass), 24 residual municipalities / 11,023 people, unchanged.

Findings, recorded in enrichment notes, not propagated to roster.jsonl (append-only; needs a future re-key/reconciliation pass):

• wv-de-3580 "Gemeinde Hille (Wasserversorgung)" — likely mis-keyed identity. hille.de never names an operating entity for its own water; its own Trinkwasserampel page states Hille is actually supplied by "WBV des Amtes Hartum," whose own site (wbv-hartum.de) shares an identical address and phone number with what hille.de publishes for its water contact. Strong evidence this company_id and a genuine WBV-Hartum roster entry are the same real-world operator keyed to the wrong domain, or a duplicate exists elsewhere on the roster. Not merged (domain-collision rule); flagged for human reconciliation.
• wv-de-1502 (WZV Woringen) — roster contradiction. Own member table (raw HTML, missed by WebFetch's summariser) lists 8 rows: 3 full members plus 5 partial members named as "Gemeinde X für den Ortsteil Y". Resolves the roster's "Benningen not a member" finding (true for the main village, false for its Benninger-Einöde exclave) but does not include Ungerhausen at all, contradicting the roster's medium-confidence claim sourced from an Ungerhausen-hosted PDF title. Flagged, not silently resolved either way.
• wv-de-3862 (WBV Usingen) — roster had self_only; operator's own Verbandsgebiet page names 3 full member towns (Neu-Anspach, Usingen, Wehrheim, with a Merzhausen/Wilhelmsdorf carve-out). 2 new links added.
• wv-de-3864 (Stadtwerke Bitburg) — analysis PDF names 2 more supply Ortsteile (Badem, Königswäldchen) beyond the roster's previously-recorded set.
• wv-de-232 (VG Loreley) and wv-de-238 (VG Kastellaun) — both upgraded from count_only to full list_found supply-area status via primary-source pages/PDFs a prior pass hadn't reached.
• wv-de-1756 (ZV Kugelberggruppe) — own Mitglieder page now lists a 7th member (Markt Biberbach) beyond the roster's seeded 6; recorded medium-confidence, not independently corroborated.
• wv-de-3825 (Stadtwerke Halver) — new TLS-cert-chain trap (same class as bodensee-wasserversorgung.de/kelsterbach.de/stadtwerke-kronach.de), worked via curl -k. Also: its purchased "WW Jubach" supply shares a name with neighbouring Stadtwerke Kierspe's own "Wasserwerk Jubach" a few km away — unresolved naming coincidence, flagged for a human.
• wv-de-3582 (Petersberg) — roster's impressum_url 404s, correct one is petersberg.de/index.html?alias=impressum_petersberg; incidental finding of 2026 news coverage of a drought-driven Trinkwassernotstand, unrelated to the schema but worth a human's attention.
• wv-de-3863 (Eppelheim) — well-earned negative: the city's own emergency-contacts table has a "Wasser (Stadtwerke Heidelberg)" row with an empty phone cell, Gas row filled in alongside it.
• New host/tooling trap: brakel.de and wzv-woringen.de/ueber-uns serve ISO-8859-15 pages that grep/ugrep silently treats as binary (zero matches) unless -a is passed — cost wasted round-trips before recognized. Worth adding to SKILL.md's Memory safety list.
• No non-suppliers found in this batch — all 45 are genuine, currently active drinking-water suppliers with real Trinkwasser content.
• No pdftotext/python3 in this environment again; both shards used the repo-root .pdftext.mjs node/zlib FlateDecode extractor, which worked cleanly where tried (VG Loreley, Gruppenwasserwerke Bornheim).

Still stale: export_csv.mjs still reads the dead legacy companies.jsonl/links.jsonl pair, not roster.jsonl/roster-links.jsonl — skipped again in favor of the live map_roster.mjs → consolidate_roster.mjs → coverage.mjs pipeline, as every recent run has done. Still worth a maintainer fixing (8+ consecutive runs now).

For next run: continue with build_enrich_batch_excl.mjs 45. 0 pending_domain. Open human follow-ups (compounding across runs, none touched this pass unless noted): (1) wv-de-20661 Gauting → Würmtal-Zweckverband re-key; (2) wv-de-20662/wv-de-3587 Gilching duplicate; (3) wv-de-3751 Heilbad Heiligenstadt likely non-supplier / EW Wasser GmbH may be the real operator; (4) export_csv.mjs pointed at dead source files; (5) wv-de-16016 Ainring and wv-de-3724 Prenzlau self_only roster flags contradicted by their own sites; (6) wv-de-3763 Oberasbach RangauWerke freshness gap; (7) stadtwerke.freilassing.de HTTP 500 — retry, don't trap yet; (8) wv-de-3800 Sinzig → Stadtwerke Rhein-Ahr AöR merger; (9) wv-de-3815 Kronach HRB conflict, blocked on stadtwerke-kronach.de TLS trap; (10) wv-de-3580 Hille likely duplicate/mis-key with WBV Hartum (new this run); (11) wv-de-1502 WZV Woringen Ungerhausen membership contradiction (new this run); (12) wv-de-3825 Halver/Kierspe "Wasserwerk Jubach" naming coincidence (new this run).