Litigation Check — Company Litigation History Screener

Screen any counterparty for litigation history before you sign a contract or onboard a client. This Actor checks a company name against US federal PACER dockets, UK case law and Poland's SAOS judgment index, and returns one row per company × jurisdiction with the case count and the top cases found — no API key, no login, no browser.

What you get

• US federal dockets by party name, not full text. CourtListener's structural party:"..." search matches the company as an actual named party in a PACER docket — not a passing mention somewhere in a document.
• UK case law, same structural guarantee. England & Wales and UK Supreme Court judgments, filtered by the official caselaw.nationalarchives.gov.uk API's own party= parameter — a genuine party-to-the-case match, not a full-text scan. (For "Tesla, Inc." that's 3 real Tesla-as-litigant cases, not the 42 you'd get from a plain keyword search.)
• Poland, honestly labeled. SAOS is a full-text search — every Polish row carries signal:"fulltext-mention" so you never mistake a keyword hit for a confirmed litigant the way US/UK rows are.
• A clean history is a result, not a null. caseCount: 0 with found: true means the check ran and came back clean — that's the due-diligence answer you're paying for, not a failed lookup.
• No API key, no login, no browser. Three keyless public HTTP endpoints, called directly.
• Runs on Apify: schedule it, monitor it, call it from the API or the MCP server, export to JSON, CSV or Excel, or push straight into your own pipeline.

How to run it

1. Click Try for free — no card needed on the free plan.
2. Paste company legal names into Companies, one per line — e.g. Tesla, Inc..
3. Pick which Jurisdictions to check: US, UK, PL (defaults to US only).
4. Press Start. Results appear in the dataset — read them in the UI, pull them from the API, or have a webhook push them onward.

Pricing

Pay-per-event: $0.005 per run start + $0.02 per completed check (one company × one jurisdiction). No monthly seat, no minimum.

A completed check is billed whether the result is dirty or clean — caseCount: 0 is a real answer, not a failure. What is not billed: a row where the source itself failed — timeout, HTTP error, a response cut off mid-read, or a payload that isn't the shape this Actor expects (e.g. a CDN error page instead of real data) — or where you asked for a jurisdiction this Actor doesn't cover. All of these come back as found: false with an error, free.

5 companies × 3 jurisdictions = 15 checks ≈ $0.305. The worst-case input (20 companies × 3 jurisdictions = 60 checks, this Actor's runtime cap) ≈ $1.21.

Set a lower run budget (maxTotalChargeUsd) than the input implies, and the Actor stops early rather than exceeding it: the pairs that were never checked simply aren't in the dataset, and the last row explains that the run halted for budget reasons — see Output.

If you watch this Actor's run log, a chargedCount number sometimes appears next to a "budget limit reached" line. That is the Apify SDK's own internal count of billed events for a single result row — typically 2 (one for result-found, one for the platform's own per-dataset-item bookkeeping event), not the number of paid rows delivered and not a dollar amount. Ignore it for cost purposes; the dataset row count and the Pricing math above are the real numbers.

Both promises above — that error/unsupported-jurisdiction rows are free, and that the worst case tops out at $1.21 — hold as long as this Actor's own Pricing tab prices only apify-actor-start and result-found. If its automatic per-dataset-row event were ever priced too, every row (including the "free" ones) would carry that extra charge; this Actor checks for that on every run, and rather than let that happen silently, it fails the run outright — no rows at all, not even an explanation row, since writing one would itself be billed at that broken price — see Output.

Input

Field	Required	What it does
companies	yes	Legal names to screen, one per line. Up to 20. Use the full legal name for the cleanest match (Tesla, Inc., not Tesla).
jurisdictions	no	Which court systems to check: US, UK, PL. Defaults to ["US"]. Up to 3. Any other code comes back as a free "unsupported jurisdiction" row instead of being rejected.
maxCasesPerRow	no	How many individual cases to list per row, 1-50 (default 20). caseCount reports the true total matched for US/PL — approximate above ~2000 matches, see Output — even when fewer are listed; for UK it can be a lower bound when partial is true (see Output).
maxConcurrency	no	How many companies to check in parallel, 1-2 (default 2). Each company's jurisdictions are always fetched in parallel with each other.

{
    "companies": ["Tesla, Inc."],
    "jurisdictions": ["US"]
}

Output

One dataset row per company × jurisdiction. A run can end early with one of five self-explanatory final rows instead of the full result set — always check sourceNote on whichever one shows up. For most of these, "early" means genuinely unchecked pairs remain (if every pair you asked for already has a row, no halt row is added even if the budget ran out on the very last one) — the one exception is the delivery-error row, noted below, where that is not knowable.

• company: "(run halted — budget exhausted)" — your run's maximum charge was reached; the remaining pairs were never checked or charged for. Raise maxTotalChargeUsd to get the full set.
• company: "(run halted — pricing misconfigured)" — result-found itself isn't billable on this run (a deployment/Pricing-tab mismatch), and this Actor refuses to give away paid rows for free rather than silently working for nothing. That one is on us, not your budget: raising maxTotalChargeUsd will not fix it — open an issue on this Actor's page. (If instead this Actor's automatic per-dataset-item event is priced — or this run's pricing tariff could not be read at all, so that price is simply unknown — there is no halt row at all: writing one would itself risk being billed at an unconfirmed price, so the run ends as FAILED with no output rows and the reason in its status message and log, not in the dataset. This should never happen on a correctly configured deployment.)
• company: "(run halted — budget check failed)" — reading the remaining budget itself failed unexpectedly (rare, usually transient). Retrying the run should work.
• company: "(run halted — delivery error)" — a push/charge call to the platform failed partway through, on a row that would have been billed. That specific pair's row may or may not have already been delivered without being billed — this Actor can't tell from here, so it stops rather than risk a duplicate or an unbilled row. This is the one halt row that can appear even when every pair technically has a row already — it errs toward over-explaining rather than silently leaving out the caveat. Usually transient; retrying the run should work.
• company: "(run halted — delivery failure)" — a plain (free, uncharged) dataset write failed unexpectedly — an error row or an unsupported-jurisdiction row never made it into the dataset. Usually transient; retrying the run should work.

This is real data from a live run against CourtListener:

{
    "company": "Tesla, Inc.",
    "jurisdiction": "US",
    "found": true,
    "caseCount": 592,
    "partial": false,
    "cases": [
        {
            "name": "Marina Faragalla v. Tesla, Inc.",
            "court": "C.D. Cal.",
            "date": "2025-07-11",
            "number": "5:25-cv-01752",
            "url": "https://www.courtlistener.com/docket/70764837/marina-faragalla-v-tesla-inc/",
            "signal": "party"
        },
        {
            "name": "Molina v. Tesla, Inc.",
            "court": "S.D. Tex.",
            "date": "2026-06-12",
            "number": "2:26-cv-00167",
            "url": "https://www.courtlistener.com/docket/73480220/molina-v-tesla-inc/",
            "signal": "party"
        }
    ],
    "sourceNote": "US federal PACER dockets via CourtListener v4, structural party-name search (party:\"Tesla, Inc.\"). 592 total docket(s) matched; showing up to 20.",
    "checkedAt": "2026-07-28T18:41:00.000Z"
}

Field	What it means
company	The company name you passed — except on the one halt row a run can end with, where it's a self-explanatory placeholder ("(run halted — budget exhausted)", "(run halted — pricing misconfigured)", "(run halted — budget check failed)", "(run halted — delivery error)", or "(run halted — delivery failure)", see above; a mis-tariffed dataset-item price, or a tariff that couldn't be read at all, produces no row at all, only a FAILED run status, see above)
jurisdiction	US, UK or PL — null on a halt row
found	Whether the check completed — true even for a clean (caseCount: 0) result. false when the source itself failed, the jurisdiction isn't supported, or (on the one possible halt row) the run stopped early
caseCount	Total cases matched. For US/PL, the source's own total regardless of pagination — exact for PL, and exact for US up to roughly 2000 matches (CourtListener switches to a cardinality-estimate count above that volume, +/-~6%; still the right number to report, just not exact at high volume — sourceNote flags it when this applies). For UK it's the count on the one feed page fetched — a lower bound when partial is true
partial	true only for UK, only when more result pages exist beyond the one fetched — meaning caseCount undercounts the true total. Always false for US/PL. A partial: true row is still billed: the check ran and returned a real (if not exhaustive) answer
cases	Up to maxCasesPerRow cases: name, court, date, number (docket/case number), url, and signal ("party" — a named party in an official docket/case title, US or UK, or "fulltext-mention" — a keyword hit, Poland only)
sourceNote	Which source answered and how to read the count
error	Present only when found is false — why the check didn't complete
checkedAt	When this Actor ran the check

Other tools we built

Need the rest of the picture?

These run on the same account, take the same shape of input and bill the same way, so they slot into an existing pipeline without new plumbing.

Actor	What it does
AI Answer Change Alert	Track whether an AI assistant's answer to a query you care about changed since last time — new sources…
AI Crawler Access Checker	Check which AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended & more) can access your website
AI Overview Citation Tracker	For each query that matters, see which sources and domains AI assistants cite in their answer — grounded…
B2B Lead Enricher	Turn a list of company websites into sales-qualified lead cards: detected tech stack, a rough revenue…
Chotot Vietnam Listings Scraper	Pull live classified ad listings (real estate, vehicles, electronics, jobs and more) straight from…

FAQ / Limitations

Is this a legal opinion? No. This is a screening tool — a fast first pass to flag counterparties worth a closer look, not a substitute for a lawyer or a paid PACER/Companies House pull.

Why only US, UK and Poland? These are the jurisdictions with a public, keyless, machine-readable search that actually works today. Brazil's DataJud API structurally cannot return party names (LGPD — the field isn't in the response at all), and the Netherlands' public case-law search isn't built to search by company name. Both are excluded rather than faked with a workaround that would silently mislead you.

Why is Poland different from US/UK? SAOS is a full-text search over judgment documents, not a structural party-name index like CourtListener or find-caselaw. A hit means the company name appears somewhere in the text — it could be the litigant, a cited third party, or a passing reference. Every Polish row is tagged signal:"fulltext-mention" for exactly this reason, and SAOS's own ingestion lags real court activity by months.

What this is NOT. It does not read case documents or judgment text, does not assess case merit or outcome likelihood, and does not cover state/provincial courts, arbitration, or jurisdictions outside US/UK/PL.

Found a wrong result, or need a jurisdiction we don't cover? Open an issue on this Actor's page.

---

Built by zinin. Questions? Telegram @timzinin.