A CRM you can run yourself. Move one pipeline first.

#!/usr/bin/env python3 """Check for a Twenty CSV import. Standard library only. Usage: python3 twenty_csv_check.py companies.csv [people.csv] Rules come from Twenty's CSV prep guide (docs.twenty.com). Read only.""" import csv, re, sys LIMIT = 10000 EMAIL = re.compile(r"^[^@\s]+@[^@\s]+\.[^@\s]+$") THOUSANDS = re.compile(r"^\d{1,3}(,\d{3})+(\.\d+)?$") def load(path): raw = open(path, "rb").read() try: text = raw.decode("utf-8-sig") except UnicodeDecodeError: print(f"{path}: FIX not UTF-8, re-save as UTF-8 CSV") sys.exit(1) return list(csv.DictReader(text.splitlines())) def col(rows, *names): for n in names: if rows and n in rows[0]: return n return None def check(path, kind): rows = load(path) problems = [] if len(rows) >= LIMIT: problems.append(f"{len(rows)} rows, the limit is {LIMIT} per file. Split it.") key = col(rows, "Domain / Domain URL", "domain") if kind == "companies" else col(rows, "email", "Emails / Primary Email") seen = {} for i, r in enumerate(rows, start=2): for name, v in r.items(): v = (v or "").strip() if v in ("true", "false"): problems.append(f"line {i}: {name} is {v}, use {v.upper()}") if THOUSANDS.match(v): problems.append(f"line {i}: {name} is {v}, remove the thousands separator") if key: v = (r.get(key) or "").strip().lower() if v: if v in seen: problems.append(f"line {i}: duplicate {key} {v} (first on line {seen[v]})") seen[v] = i if kind == "companies" and key and r.get(key) and not r[key].startswith("https://"): problems.append(f"line {i}: domain {r[key]} should look like https://domain.com") if kind == "people": e = (r.get("email") or "").strip() if e and not EMAIL.match(e): problems.append(f"line {i}: email {e} is not a valid address") d = (r.get("companyDomain") or "").strip() if d and not d.startswith("https://"): problems.append(f"line {i}: companyDomain {d} should look like https://domain.com") return rows, problems def main(): args = sys.argv[1:] if not args: sys.exit(__doc__) crows, cprob = check(args[0], "companies") allprob = {args[0]: cprob} if len(args) > 1: prows, pprob = check(args[1], "people") known = {(r.get("Domain / Domain URL") or "").strip().lower() for r in crows} for i, r in enumerate(prows, start=2): d = (r.get("companyDomain") or "").strip().lower() if d and d not in known: pprob.append(f"line {i}: companyDomain {d} is not in {args[0]}. Import companies first.") allprob[args[1]] = pprob bad = 0 for path, probs in allprob.items(): print(f"{path}: {len(probs)} problem(s)") for p in probs: print(" " + p) bad += len(probs) print("Ready to import." if not bad else "Fix the lines above, then import companies first, people second.") sys.exit(1 if bad else 0) main()