add dolibarr-tva-reconciliation, dolibarr-recurring-templates, dolibarr-data-snapshot

V3 bundle — three sibling skills under .claude/skills/, all read-only,
all depending on the dolibarr base skill.

dolibarr-tva-reconciliation:
- tva-by-month.sh: HT + TVA grouped by (year-month × tva_tx), ready
  for CA3 / CA12 transcription.
- tva-line-detail.sh: per-line audit trail with country-based bucket
  assignment (A1 domestic / A4 intra-UE autoliquidation / E2 export
  hors UE). Documents the French TVA mental model.
- Today every Arcodange line is E2 (KissMetrics, US, autoliquidation
  259-1° CGI). The skill scales for the day a French B2B is invoiced.

dolibarr-recurring-templates:
- list-templates.sh: probes /invoices/templates/{id} since there's no
  list endpoint. Stops after 5 consecutive empty responses.
- inspect-template.sh: full audit per template, with health checks.
- Surfaces that the "Kiss Metrics Invoice" template has frequency=0
  and nb_gen_done=0 — it is NOT auto-firing. Every KM invoice today
  was manually duplicated. Cohort-review implication: the deferred
  9-month cycle depends on Gabriel clicking "Generate" each month,
  not on a Dolibarr cron.

dolibarr-data-snapshot:
- snapshot.sh: bundles every read endpoint the dolibarr-* family uses
  into one JSON with a content_hash (sha256 of data only, excluding
  timestamp — so identical state hashes identically across runs).
- Use cases: cohort evidence packs, drift detection, archival before
  a known-risky UI change.
- V1 baseline summary captured at examples/snapshot-summary.txt
  (the ~246 KB snapshot file itself is intentionally not committed).

Also extends dolibarr/SKILL.md endpoint catalogue with
/invoices/templates/{id} (and its no-list-endpoint quirk + the
id-null sentinel for missing ids), plus links to the three new
sibling skills.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

.claude/skills/dolibarr-tva-reconciliation/scripts/tva-line-detail.sh

@@ -0,0 +1,98 @@
+#!/usr/bin/env bash
+# Per-line TVA breakdown with the country classification needed to choose
+# the right CA3 line (E2 export hors UE vs A4 autoliquidation intra-UE vs
+# domestic A1).
+#
+# Usage:
+#   tva-line-detail.sh [--year YYYY] [--since YYYY-MM-DD] [--until YYYY-MM-DD]
+#
+# Pulls each invoice's lines + the thirdparty's country_code to assign each
+# line to a CA3 bucket. Useful as the audit trail behind tva-by-month.sh.
+
+set -euo pipefail
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+DOL_CURL="${SCRIPT_DIR}/../../dolibarr/scripts/dol-curl.sh"
+
+SINCE=""; UNTIL=""; YEAR=""
+while [[ $# -gt 0 ]]; do
+  case "$1" in
+    --since) SINCE="$2"; shift 2 ;;
+    --until) UNTIL="$2"; shift 2 ;;
+    --year)  YEAR="$2"; SINCE="$2-01-01"; UNTIL="$2-12-31"; shift 2 ;;
+    -h|--help) sed -n '2,12p' "$0" | sed 's/^# \{0,1\}//'; exit 0 ;;
+    *) echo "tva-line-detail.sh: unknown arg: $1" >&2; exit 2 ;;
+  esac
+done
+
+WORK="$(mktemp -d -t tvaline.XXXXXX)"
+trap 'rm -rf "${WORK}"' EXIT
+
+"${DOL_CURL}" '/invoices?limit=500&sortfield=t.datef&sortorder=ASC' > "${WORK}/inv.json"
+
+# Per-invoice detail + per-thirdparty country
+IDS=$(python3 -c "import json,sys; print(' '.join(str(r['id']) for r in json.load(open(sys.argv[1]))))" "${WORK}/inv.json")
+mkdir -p "${WORK}/inv" "${WORK}/tp"
+SOCIDS=$(python3 -c "import json,sys; print(' '.join(sorted({str(r.get('socid')) for r in json.load(open(sys.argv[1])) if r.get('socid')})))" "${WORK}/inv.json")
+for id in ${IDS};    do "${DOL_CURL}" "/invoices/${id}"     > "${WORK}/inv/${id}.json"; done
+for s  in ${SOCIDS}; do "${DOL_CURL}" "/thirdparties/${s}" > "${WORK}/tp/${s}.json"; done
+
+python3 - "${WORK}" "${SINCE}" "${UNTIL}" "${YEAR}" <<'PY'
+import json, sys, os, datetime
+work, since, until, year_arg = sys.argv[1:5]
+
+# EU member states (excl. FR — FR lines are "domestic")
+EU = set("AT BE BG HR CY CZ DK EE FI DE GR HU IE IT LV LT LU MT NL PL PT RO SK SI ES SE".split())
+
+tp = {}
+for fn in os.listdir(os.path.join(work,"tp")):
+    try: d = json.load(open(os.path.join(work,"tp",fn)))
+    except json.JSONDecodeError: continue
+    tp[fn[:-len(".json")]] = d.get("country_code") or ""
+
+def bucket(country, tx):
+    if tx and tx > 0: return f"A1 (domestic {tx}%)"
+    if country == "FR": return "A1 (domestic, 0%? check)"
+    if country in EU:   return "A4 (autoliquidation intra-UE)"
+    if country:         return "E2 (export hors UE)"
+    return "(unknown country)"
+
+def in_window(ts):
+    if not ts: return False
+    d = datetime.date.fromtimestamp(int(ts))
+    if since and d < datetime.date.fromisoformat(since): return False
+    if until and d > datetime.date.fromisoformat(until): return False
+    return True
+
+print(f"{'date':<10}  {'invoice':<24} {'client':<15} {'cnty':<4} {'tx':>5}  {'HT':>10}  {'TVA':>8}  CA3 bucket")
+print("-" * 110)
+all_lines = []
+for fn in sorted(os.listdir(os.path.join(work,"inv"))):
+    try: inv = json.load(open(os.path.join(work,"inv",fn)))
+    except json.JSONDecodeError: continue
+    ts = int(inv.get("date") or 0)
+    if not in_window(ts): continue
+    dt = datetime.date.fromtimestamp(ts)
+    sid = str(inv.get("socid") or "")
+    cnty = tp.get(sid, "")
+    client = "tp-"+sid
+    for line in inv.get("lines") or []:
+        tx  = round(float(line.get("tva_tx") or 0), 4)
+        ht  = float(line.get("total_ht")  or 0)
+        tva = float(line.get("total_tva") or 0)
+        buck = bucket(cnty, tx)
+        all_lines.append((dt, inv["ref"], client, cnty, tx, ht, tva, buck))
+        print(f"{dt}  {inv['ref']:<24} {client:<15} {cnty:<4} {tx:>5.2f}  {ht:>10.2f}  {tva:>8.2f}  {buck}")
+
+print("-" * 110)
+# Summary per bucket
+import collections
+buckets = collections.defaultdict(lambda: {"ht":0.0,"tva":0.0,"n":0})
+for r in all_lines:
+    b = r[7]
+    buckets[b]["ht"] += r[5]; buckets[b]["tva"] += r[6]; buckets[b]["n"] += 1
+print()
+print("# Aggregated by CA3 bucket:")
+for b, s in sorted(buckets.items()):
+    print(f"  {b:<35} count={s['n']:>3}  HT={s['ht']:>10.2f}  TVA={s['tva']:>8.2f}")
+PY