MORVS

View saved version →

---
name: compliance-screening
description: This skill should be used when a user asks to screen an organization against public sanctions or debarment records, compare counterparties, or review regulatory and complaint signals with coverage and identity caveats.
---

# Compliance screening

**Lead with this limit:** MORVS screens are research aids with partial, variable coverage. A non-match, low score, or failed experimental tool is not sanctions clearance, compliance approval, or a legal conclusion. Verify identity and current status in the official source before any action.

## Workflow

1. Keep scope to organizations and institution-level facts; do not reproduce personal identifiers, addresses, or complaint narratives containing PII. Resolve legal names and aliases with `resolve` using `{"q":"Organization name"}` or a known ticker or CIK. Record unresolved name collisions.
2. For a single organization, call `sanctions_check` with `{"name":"Exact legal name"}` and `debarments_check` with the same name. The sanctions tool searches a partial local OFAC SDN index by substring, including indexed aliases. Debarments covers local SAM Exclusions and HHS OIG LEIE records. Treat hits as possible matches pending identifier and current official-list review.
3. When broader context is requested, call `screen_all` with `{"name":"Exact legal name"}` if available. It is experimental; its 0–100 score and per-source counts are indicative local-index outputs. For up to 50 names call `screen_batch` with `{"names":["Name A","Name B"]}`. For two entities call `compare_entities` with `{"a":"Name A","b":"Name B"}`. Describe any ranking as a MORVS heuristic, not a determination of risk or wrongdoing.
4. Use `doj_by_company` with `{"name":"Exact legal name"}` to inspect DOJ press-release mentions and `cfpb_by_company` with the same argument to inspect a partial CFPB complaint index. A press mention need not be an enforcement action; a complaint is a consumer report, not a proven violation.
5. Inspect MORVS coverage and call `dataset_freshness` with `{}` for relevant federal hub rows. Verify material hits and negatives against current official [OFAC](https://ofac.treasury.gov/sanctions-list-service), [SAM.gov](https://sam.gov/content/exclusions), [HHS OIG](https://exclusions.oig.hhs.gov/), [DOJ](https://www.justice.gov/news), or [CFPB](https://www.consumerfinance.gov/data-research/consumer-complaints/) records as applicable. Preserve record IDs, dates, list version or access date, and exact source URLs. Say when current official verification was unavailable.
6. Put the caveat before the results, then provide a compact table. Report no-hit only as “no match in the checked local index on the access date.” Avoid accusations and definitive clearance language.

| Checked on | Organization and identifier | Source checked | Local result | Official record/date/URL | Match and coverage limit |
|---|---|---|---|---|---|
| UTC date | Legal name plus verified ID | Agency and dataset | Hit, no local hit, or unavailable | Exact record or list URL | Alias/identity and freshness status |

## Worked examples

**One vendor.** For “Screen Microsoft for sanctions and exclusions”, call `resolve` with `{"ticker":"MSFT"}`, `sanctions_check` with `{"name":"Microsoft Corporation"}`, and `debarments_check` with the same name. Confirm any candidate match at current OFAC, SAM.gov, or HHS OIG official records before reporting it. If there is no local hit, report only the local-index result and its coverage limit.

**Two counterparties.** For “Compare Pfizer and Merck”, call `compare_entities` with `{"a":"Pfizer","b":"Merck"}`; then call `sanctions_check` and `debarments_check` separately for each verified legal name. Use `doj_by_company` and `cfpb_by_company` only for relevant context, opening the original records. Explain that the comparison score is a heuristic and that name ambiguity or source lag can change it.

View saved version →

---
name: dataset-finder
description: This skill should be used when a user asks to find a MORVS or Voidly public dataset, identify a keyless JSON endpoint, check dataset dates and coverage, or understand source attribution and reuse terms.
---

# Dataset finder

The public [MORVS and Voidly dataset catalog](https://morvs.ai/voidly/datasets.json) is a static, keyless JSON document. Fetch it when answering; do not rely on a remembered entry or count. It is distinct from MORVS's federal data hub and its MCP tools.

## Workflow

1. Fetch the catalog through an available read-only HTTPS tool, such as Claude Code WebFetch or Codex web.run. Search entries by topic, name, ID, and description. Do not invent a catalog search MCP tool; none appears in the 2026-10-09 live 73-tool list.
2. For each relevant entry, record the landing page, JSON endpoint, extra endpoints, description, counts, build date or observed-at time, and stated license or rights. Follow the entry's endpoint and source/provenance links before claiming current coverage. A catalog count is a dated observation, not a live promise.
3. Read license and rights at both catalog and entry level. For entries derived from third-party material, follow upstream terms; a compilation license does not automatically cover raw source data. State attribution and commercial-reuse limits only when verified at the linked terms.
4. For a federal data hub source found through an entity, call `resolve` with a ticker, CIK, or name to inspect the available-in map, then call `dataset_freshness` with `{}` for source-specific cadence, last successful observation, and indexed-content watermark. Do not apply that freshness result to the static catalog entries unless the dataset is actually represented in the returned row.
5. Return only institution-level or aggregate facts. Avoid PII, unsupported accusations, or treating data absence as evidence of no activity.

| Dataset | Best use | Landing / JSON endpoint | Data date and coverage | Rights / upstream terms | Checked on |
|---|---|---|---|---|---|
| Catalog name | Plain-language use | Both clickable links | Explicit date semantics and gaps | Exact catalog or source terms | Access date |

## Worked examples

**Banned-books data.** For “Find a reusable banned-books JSON source”, use WebFetch or web.run to fetch [the catalog](https://morvs.ai/voidly/datasets.json), select its Verboten entry, and open its listed landing page and endpoint. Read the entry's license and source links. Report the current endpoint, its date/coverage, and the stated terms with citations; do not copy a remembered catalog count as live coverage. No MORVS MCP catalog tool is needed.

**Company-linked federal dataset.** For “Which MORVS federal datasets mention Pfizer?”, call `resolve` with `{"ticker":"PFE"}`. Use its available-in map and links to choose datasets, then call `dataset_freshness` with `{}` and match returned dataset rows. If the user also wants a static Voidly dataset, fetch [the separate catalog](https://morvs.ai/voidly/datasets.json) and check that entry's own endpoint and rights. Keep the two coverage systems separate.

View saved version →

---
name: market-signals
description: This skill should be used when a user asks about a public company's insider activity, FINRA short interest or short-sale volume, SEC fails to deliver, or a possible squeeze signal with dated source evidence.
---

# Market signals

Treat all signal rankings as research leads, not trading recommendations or proof of manipulation. FINRA short-sale volume is not short interest; a fail to deliver is not by itself evidence of naked shorting. Form 4 records executed transactions; Form 144 records proposed sales.

## Workflow

1. Confirm the issuer and ticker with `resolve`, for example `{"ticker":"NVDA"}`. Check share classes and security identifiers before joining datasets.
2. For insider activity, call `company_insider_pulse` with `{"ticker_or_cik":"NVDA","days":90}`. Use `cluster_buys` with `{"days":90,"min_insiders":3}` or `cluster_sells` with `{"days":30,"min_insiders":4}` for discovery; then inspect the underlying filings for the named issuer. Distinguish transaction code, transaction date, filing date, executed trades, and planned sales. Summarize at issuer level without repeating individuals' personal details.
3. For FINRA, call `short_interest_search` with `{"q":"NVDA"}` for settlement-date short positions and days to cover. Call `short_volume_search` with `{"q":"NVDA"}` for the separate rolling short-sale volume series. Do not add or equate the two measures. Respect [FINRA source terms](https://www.finra.org/terms-of-use).
4. Call `fails_to_deliver_search` with `{"q":"NVDA"}` for SEC fail balances and days with fails. Do not sum daily outstanding balances into a claim of unique unsettled shares without a valid method.
5. If asked for candidates, call `short_squeeze_candidates` with `{"min_score":70,"limit":20}` or `short_interest_squeeze` with `{}`. The MORVS pressure score combines inputs with a heuristic. Verify each candidate's ticker, settlement date, data window, liquidity, corporate actions, and source rows before interpretation.
6. Call `dataset_freshness` with `{}` and match relevant dataset rows. State the latest available FINRA settlement, trading, SEC reporting, and MORVS observation dates separately. Cite exact SEC filings and data records when supplied; if only an aggregate endpoint is returned, label it an aggregate and link it.

| Source date or window | Issuer/security | Measure or event | Value and units | Primary record URL | Interpretation limit |
|---|---|---|---|---|---|
| Settlement, trade, or filing date | Ticker and identifier | Exact metric or filing | As returned, no invented precision | Link to record | Lag, coverage, and metric caveat |

## Worked examples

**One ticker's short-selling picture.** For “What is happening with NVDA shorting?”, call `resolve` with `{"ticker":"NVDA"}`, `short_interest_search` with `{"q":"NVDA"}`, `short_volume_search` with `{"q":"NVDA"}`, and `fails_to_deliver_search` with `{"q":"NVDA"}`. Report each source's date and units in separate rows. Do not interpret a high short-volume percentage as a high open short position.

**Insider and squeeze discovery.** For “Find recent insider buys and possible squeeze candidates”, call `cluster_buys` with `{"days":90,"min_insiders":3}` and `short_squeeze_candidates` with `{"min_score":70,"limit":20}`. For any overlapping ticker, call `company_insider_pulse` with that ticker and `short_interest_search` plus `fails_to_deliver_search` with the same ticker. Cite the underlying filings and source windows; call it a screened lead, not a buy signal.

View saved version →

---
name: public-records-research
description: This skill should be used when a user asks to research a company in public records, assemble an entity timeline, inspect a cross-source dossier, or find material events with dated source citations.
---

# Public records research

Build a source-linked, institution-level account from MORVS public datasets. Treat a name match as a lead until an identifier and underlying record support it.

## Workflow

1. Define the organization, jurisdiction, period, and question. Avoid collecting or repeating personal identifiers, residential addresses, or other PII. Report institution-level facts.
2. Call `resolve` with exactly one known identifier when possible: `{"ticker":"PFE"}`, `{"cik":"0000078003"}`, or `{"q":"Pfizer"}`. Inspect canonical name, aliases, identifiers, available sources, and source links. Disambiguate similar names before proceeding.
3. Select the narrowest useful tool:
   - `entity_timeline` with `{"key":"PFE","days":180,"limit":100}` for a dated cross-source event sequence. The key accepts ticker, CIK, UEI, LEI, DUNS, or numeric entity ID; maximum days is 730.
   - `entity_material_events` with `{"key":"PFE","days":180}` for the high-signal subset. Its importance threshold is a MORVS selection rule, not an official finding.
   - `dossier` with `{"ticker":"PFE"}` or `{"name":"Pfizer"}` for the cross-topic overview.
   - `company_profile` with `{"ticker_or_cik":"PFE"}` for an SEC-filing company profile. Do not use its absence to infer that a private organization has no records.
4. Check the returned period, per-source coverage, record dates, and source links. Call `dataset_freshness` with `{}` if using a federal hub dataset and match its row to the dataset; distinguish source-observation time from indexed-content watermark. Mark unknown or stale coverage plainly. A missing hit is not evidence of no activity.
5. Open the primary record URL for material claims when available. Distinguish a filing, complaint, press release, recall, contract, and agency finding. State only what the source establishes. Never turn a mention, allegation, or automated flag into an accusation.
6. Return a short answer followed by a record table. Give access date and source date; omit a claim when no verifiable record URL is available.

| Event or fact date | Institution-level finding | Dataset and source | Record ID | Primary record URL | Coverage or freshness |
|---|---|---|---|---|---|
| Use the record's date | State only the supported fact | Name the agency and MORVS index | If supplied | Link to the exact record | State the source window and any gap |

## Worked examples

**A public company's recent regulatory history.** For “What happened at Pfizer over the last six months?”, call `resolve` with `{"ticker":"PFE"}`, then `entity_timeline` with `{"key":"PFE","days":180,"limit":100}` and `entity_material_events` with `{"key":"PFE","days":180}`. Use `company_profile` with `{"ticker_or_cik":"PFE"}` only if financial and SEC context matters. Reconcile duplicates by record ID and date; cite each original filing or agency record. Do not infer misconduct from a high importance score.

**An ambiguous organization name.** For “Research Acme Health”, call `resolve` with `{"q":"Acme Health"}`. Present candidate organizations and identifiers if more than one appears. After a verified match, call `dossier` with the exact organization name and `entity_timeline` with its returned identifier. If no identifier resolves, give only clearly attributed search leads and say the match is unconfirmed.