# LiquiLens — bank and lender research [Verified] [Health: Active]

**Category:** 💰 Finance & Fintech  
**Repository:** https://github.com/beepboop2025/liquilens-mcp  
**GitHub Stars:** 0  
**Views:** 6  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/liquilens-the-failure-radar

## Description
Source-dated bank, NBFC and MFI research with 24 read-only tools on the public MCP endpoint. Review NPA disclosures, institution evidence, funding transmission and historical diagnostics with citations, coverage gaps and methodological limits. Historical results retain their construction-PIT and eligibility labels; they do not establish validated forecasts, ratings or trade-execution authority.

## Tools
Capabilities this server exposes over MCP:

- **bank_asset_quality_review** — Cited NPA changes, PCR definitions and capital scope for banks/SFBs/UCBs, plus separately dated provisional business updates when accepted. Keep each period and provisional status explicit. No score, credit or execution authority.
- **bank_npa_reconciliation** — Reconcile supplied NPA movements; keep recoveries/write-offs/sales/upgrades distinct. Arithmetic, not input authentication.
- **banking_specialisation_coverage** — Bank/SFB/UCB coverage with stale and missing evidence. Research, not a census or rating.
- **corporate_transmission_board** — Answers ONE question: is US funding stress reaching nonfinancial firms? Use household_credit_board for households. CP, bank credit lines and real-economy confirmation; cannot_see preserves gaps. Seiche context never enters regime or transmission. Display-only; no institution score or tier. full:true adds methods.
- **crypto_exposure_board** — Read the compact cited bank/crypto exposure register: disclosed quantum, stablecoin and venue links, exposure channels, Undertow run-risk cross-read and compound flags. Detailed disclosures stay on the public REST detail route. Takes no arguments; display-only, never an institution score or investment advice.
- **crypto_regime_board** — Read the compact BTC/ETH Crypto Regime board: current change-point state, log-SR statistic, last shift, observation count and the display-only cross-read against disclosed bank exposure. Time-series paths stay on the public REST detail route. Takes no arguments; feeds no institution score and is not investment advice.
- **evidence_europe** — Read seven cited European retrospective case files through unrecalibrated Indian lenses. Named case studies, not a cohort: no European recall percentage to quote. Omit slug for summaries, or supply a returned slug for a full quarterly case file.
- **evidence_india** — Read the 48-institution Indian construction-PIT crisis diagnostic, including misses and false alarms. Uses publication clocks where present and a +60-day filing-lag proxy otherwise; overwritten amendments are unreconstructable. evidence_institution returns a full replay.
- **evidence_institution** — Read a sourced Indian construction-PIT quarterly replay, lens alert timing and hit/miss/false-alarm verdict. Use evidence_india for exact slugs. Filing-availability proxies do not establish a validated backtest.
- **evidence_markets** — Historical evidence: India (48-institution construction-PIT diagnostic), US (industry-wide current-amended vintage), Europe (named cases, no cohort claim). Start here, then evidence_india, evidence_us or evidence_europe for details.
- **evidence_us** — Read US bank current-amended-vintage diagnostics and named replays, including misses. Not a validated point-in-time backtest. Report the quarterly top-decile budgeted watchlist beside recall, uncertainty and lead times; it is not an individual-bank verdict. include_trajectories:true adds histories.
- **failure_radar_board** — Vetted Indian lender PDs, PCA/SAF and funding. market_dd requires tier authority; diagnostic clocks/refusals stay in market_dd_evidence. Slugs for failure_radar_institution. Research, not a rating.
- **failure_radar_institution** — Read one Indian lender dossier: quarterly PDs and drivers, PCA/SAF headroom, funding, forensic and listed-market evidence. Discover slugs with failure_radar_board. Uncovered institutions remain absent, never scored from memory.
- **forward_odds** — Read counted Markov transitions and empirical 5/20-day reach odds. Odds are withheld until 60 observed days; counts stay visible and absence is never precision. Historical frequencies are not calibrated forecasts. Display-only, never a state driver.
- **household_credit_board** — Answers ONE question: is stress reaching US household credit? Use corporate_transmission_board for firms. Delinquency, charge-offs, revolving credit and unscored debt-service context; gaps stay in cannot_see. Display-only; no institution score or watchlist tier. full:true adds methods and histories.
- **institution_evidence_monitor** — Bank/NBFC evidence, gaps and review alerts. slug: one record; slugs: watchlist; neither: all tracked. End-of-day as_of. No scoring or delivery authority.
- **institution_research_coverage** — Search Indian dossiers or RBI rows. Overlapping populations; registry presence grants no financial coverage or score.
- **institution_review_packet** — Cited human review packet for an exact slug or unique normalized full name. No fuzzy match or new score; stale/dark gaps retained. Content hash detects alteration, not attestation or publication-time proof.
- **latest_article** — Read today's exact full-text LiquiLens editorial: current institution-risk analysis when the evidence moved, or a labelled historical replay when it did not. Returns the canonical headline, dek, Markdown, evidence clock and passing publication receipt; quote it without regenerating facts.
- **rbi_supervisory_tape** — Read the latest RBI enforcement and supervisory actions — monetary penalties, licence cancellations and supervisory directions — parsed from rbi.org.in press releases, newest first, each item carrying the RBI's own source URL. Takes no arguments. When no tape snapshot exists on the deployment, says so in a stated note instead of returning silent emptiness.
- **research_network** — Route to Palimpsest sources, Seiche funding, Undertow exit liquidity and NarcoScope data. Set limit for full metadata. Rights remain attached; no identity, score or eligibility changes.
- **stablecoin_rails_board** — Paged peg/run evidence; partial data is never CALM. Follow pagination.next_offset. Full receipts on REST. Display-only, not investment advice.
- **universe_search** — Search the RBI NBFC register by name/CIN substring, classification or regulation layer. Returns official register rows. Supply q or a filter; a filter without q browses that slice. Registration does not establish creditworthiness.
- **verify_published_record** — Verify a ledger stream by recomputing hash-chain commitments and checking enabled signatures and anchors. Returns the actual verifier verdict, including failures. Verifies record integrity, not economic accuracy.

## Claude Desktop Quick Installation
Remote MCP endpoint (confidence: high). Install path detected from listing signals. Add as a URL/SSE server in your client:

```json
"mcpServers": {
  "liquilens-bank-and-lender-research": {
    "url": "https://api.liquilens.in/mcp"
  }
}
```

## Documentation

## What LiquiLens — bank and lender research does

LiquiLens — bank and lender research provides a public, read-only MCP interface for researching banks and lenders from cited records. Its coverage includes Indian banks, small finance banks, urban cooperative banks, NBFCs, MFIs, RBI register rows, supervisory actions, and selected US and European historical evidence. The service keeps reporting periods, source dates, provisional disclosures, eligibility labels, missing evidence, and methodological limits visible.

The tool set supports several distinct research tasks. Asset-quality tools review NPA changes, provisioning coverage ratio definitions, capital scope, and supplied movement reconciliations. Coverage and evidence tools locate institutions, identify stale or missing records, and produce sourced review packets or institution replays. Failure-radar tools expose selected lender PDs, PCA or SAF status, funding information, and supporting forensic or listed-market evidence where coverage exists.

Additional boards examine whether stress is reaching US companies or households, while separate tools read crypto exposure, stablecoin and peg/run evidence, and crypto-regime indicators. RBI supervisory actions are returned with the regulator's source URL. Historical-market evidence covers Indian construction-PIT diagnostics, US current-amended-vintage analysis, and named European cases; these outputs do not establish validated backtests or forecasts.

## How it works

LiquiLens — bank and lender research is served at `https://api.liquilens.in/mcp` using Streamable HTTP. An MCP client sends research requests to the hosted endpoint, and the tools return structured results with citations or source references where available. Some tools accept an institution slug, search term, filter, pagination offset, or other documented argument; others are fixed display boards with no arguments.

The service separates what is observed from what is unavailable. Fields such as `cannot_see`, stale evidence, missing disclosures, and uncovered institutions remain explicit rather than being treated as reassuring results. Historical outputs retain their construction-PIT, current-amended-vintage, eligibility, and filing-availability labels. Reconciliation tools perform arithmetic on supplied inputs but do not authenticate those inputs. Ledger verification checks hash-chain commitments, signatures, and anchors, but not the economic truth of a record.

## Setup and configuration

Connect an MCP client to the hosted endpoint; no local server, API key, or package installation is required for the public read-only tools. Claude Code can be configured with:

```sh
claude mcp add --transport http liquilens https://api.liquilens.in/mcp
```

For Codex, the documented command is:

```sh
codex mcp add liquilens --url https://api.liquilens.in/mcp
```

Cursor users can merge the remote MCP configuration into a project `.cursor/mcp.json` or the user-level `~/.cursor/mcp.json`. The supplied Cursor configuration also includes a companion Seiche funding endpoint, so existing entries should be preserved when merging. Claude.ai or ChatGPT may use a custom connector or MCP app where the workspace supports that feature. Fair-use limits and model-provider charges may apply.

## Tools and capabilities

The public inventory contains 24 read-only tools, including:

- `bank_asset_quality_review` and `bank_npa_reconciliation` for dated NPA research and supplied-movement arithmetic.
- `institution_research_coverage`, `institution_evidence_monitor`, `institution_review_packet`, and `universe_search` for discovery and coverage checks.
- `failure_radar_board` and `failure_radar_institution` for covered Indian lender risk dossiers.
- `evidence_markets`, `evidence_india`, `evidence_us`, `evidence_europe`, and `evidence_institution` for historical diagnostics and replays.
- `corporate_transmission_board` and `household_credit_board` for US real-economy transmission questions.
- `rbi_supervisory_tape`, `latest_article`, `research_network`, and `verify_published_record` for supervisory, editorial, source-routing, and integrity tasks.

The remaining boards address crypto exposure, crypto regime conditions, stablecoin rails, and forward historical odds. These are display or research outputs, not institution scores, investment advice, calibrated forecasts, or execution controls.

## Limitations and notes

LiquiLens — bank and lender research does not provide credit ratings, validated forecasts, trade execution authority, or a guarantee of complete market coverage. Registration in a regulator's register does not establish creditworthiness, and absence from a covered list does not justify scoring an institution from memory. Public data may be stale, incomplete, provisional, or impossible to reconstruct after amendments are overwritten.

Some historical diagnostics use filing-lag proxies or current-amended data and therefore are not validated point-in-time backtests. Forward odds are withheld until the stated observation threshold and remain historical frequencies rather than calibrated predictions. Display boards can preserve gaps in `cannot_see`, and paged results must be followed through their returned pagination links. Research calls do not execute trades.

_Full upstream README: https://allmcps.com/mcp/liquilens-the-failure-radar/readme_