Build a Claude and Quiver Market Research Alert System

Help me connect this Claude Code project to Quiver Quant using my available API credentials or MCP configuration.

First, inspect the available integration method and explain what configuration is required. Then create the local project structure needed to query Quiver data.

Do not claim the connection works until you have run a small test query. For the test, request the most recent available insider transactions and show the filing date, transaction date, ticker, insider name, title, transaction type, share count, and reported value when those fields are available.

If a required endpoint, package, permission, or field is unavailable, stop and tell me exactly what is missing.

Using only the connected Quiver data, show me the 20 most recently filed corporate insider transactions.

For every row, include:
1. Company and ticker
2. Insider name and title
3. Purchase, sale, option exercise, or other transaction classification
4. Transaction date
5. Filing date
6. Share count
7. Reported dollar value
8. The Quiver dataset or record used

Sort by filing date, newest first. Clearly label missing fields. Do not fill gaps with assumptions or general web results.

Using Quiver insider trading data, analyze whether insiders at [TICKER] are buying or selling.

Use the most recent available records and separate:
1. Open market purchases
2. Open market sales
3. Option exercises
4. Grants, vesting, or compensation related transactions
5. Other transaction types

Show the transaction date, filing date, insider name, title, shares, price, and reported dollar value when available.

Then summarize:
1. The number of distinct buyers and sellers
2. Total reported purchase and sale values
3. Whether activity appears concentrated among senior executives
4. Whether several records were filed on the same date or have similar values
5. Any reason the raw totals could be misleading

Do not interpret insider activity as a recommendation to buy or sell.

Using Quiver insider trading data, find every company where three or more distinct insiders made open market purchases during the last 60 days.

Requirements:
1. Include only open market purchases.
2. Exclude option exercises, grants, awards, vesting transactions, and sales.
3. Group results by company and ticker.
4. Rank first by the number of distinct buyers, then by total reported dollar value.
5. Show every buyer's name, title, transaction date, filing date, shares purchased, price, and reported value when available.
6. Flag transactions that appear to be part of a coordinated batch or compensation process.
7. Explain each batch flag using observable evidence, such as identical filing dates, similar transaction values, repeated titles, or tightly grouped transaction dates.
8. Separate potentially independent clusters from likely administrative batches.
9. State the latest filing date included in the analysis.

Do not describe a cluster as bullish without discussing the transaction sizes and batch risk.

Stress test the insider purchase cluster for [TICKER].

Determine whether the purchases look like independent open market decisions or a coordinated administrative batch. Compare transaction dates, filing dates, reported values, share counts, insider titles, and transaction codes.

Present:
1. Evidence supporting an independent cluster
2. Evidence supporting a batch or compensation explanation
3. The largest and smallest purchase
4. Purchases by senior officers or directors
5. Any outlier transaction that differs materially from the rest
6. A conclusion labeled strong cluster, mixed evidence, or likely batch

Base the conclusion only on available records and state any missing information.

Using Quiver congressional trading data, show the most recently filed congressional trades.

For each record include:
1. Member name
2. Chamber and party, if available
3. Ticker or asset
4. Purchase or sale classification
5. Reported value range
6. Trade date
7. Filing date
8. Days between the trade and filing

Sort by filing date, newest first. Add a warning to any record where the filing delay makes it unsuitable as a real time signal.

Give me the full political exposure profile for [COMPANY OR TICKER] using Quiver data.

Analyze:
1. Congressional purchases and sales
2. Federal contract awards
3. Lobbying spending and disclosures

For congressional activity, show trade dates and filing dates separately.
For contracts, summarize reported award values, agencies, and timing when available.
For lobbying, show spending and the specific issue text from individual disclosures when available.

Finish with:
1. Areas where political or government exposure appears to be increasing
2. Areas where the evidence is mixed
3. Important reporting delays or missing fields
4. Questions that require further company research

Do not treat lobbying, contracts, or congressional trades as proof of future revenue or stock performance.

Give me a complete ownership and sentiment profile for [TICKER] using connected Quiver data.

Include:
1. Insider purchases and sales by quarter
2. Net reported insider activity by quarter
3. Institutional ownership
4. Top institutional holders
5. Material reported increases or reductions by major holders
6. Any disagreement between insider behavior and institutional positioning

Distinguish confirmed data from interpretation. If insiders are selling while institutions are split, describe that as disagreement rather than clear directional momentum.

Using Quiver congressional and corporate insider data from the last 90 days, identify tickers traded by fewer than five distinct people where activity appears to be increasing.

Compare those names with tickers that already appear crowded.

You must define increasing activity using measurable fields, such as:
1. More distinct participants in the recent period
2. Higher transaction frequency
3. Higher reported dollar value
4. A shift from selling to buying
5. Shorter gaps between relevant transactions

For each result, show the underlying records and explain which measurements caused it to qualify. Rank by signal quality, but do not equate the ranking with expected returns.

From the results above, rank the stocks that deserve the most additional research based only on the congressional and insider patterns.

For each stock provide:
1. The observed pattern
2. Why it may be notable
3. The strongest alternative explanation
4. Filing delay concerns
5. The next fundamental question to investigate

Label the ranking as a research priority list, not a buy list.

Create an alert specification for Quiver corporate insider filings.

Alert when all of the following are true:
1. The transaction is an open market purchase.
2. The ticker is [ANY TICKER OR WATCHLIST].
3. The reported purchase value is at least [MINIMUM VALUE].
4. The insider title matches [TITLE FILTER OR ANY].
5. The filing has not already generated an alert.

Increase priority when:
1. Two or more distinct insiders buy the same ticker within [NUMBER] days.
2. A chief executive, chief financial officer, director, or other selected senior role participates.
3. The purchase is materially larger than the insider's other recent purchases.

Reduce priority or flag for review when:
1. Several transactions have nearly identical values.
2. Many records share the same filing date.
3. The activity appears compensation related.
4. Required transaction fields are missing.

Each alert must include the ticker, company, insider, title, transaction date, filing date, shares, price, reported value, cluster count, and batch warning.

Build a Telegram alert service for the connected Quiver data source.

The service must:
1. Check for newly available corporate insider filings.
2. Apply the alert specification below.
3. Send matching records to my Telegram bot.
4. Store a stable filing or transaction identifier so the same record is never sent twice.
5. Include transaction date and filing date in every alert.
6. Separate open market purchases from option exercises, grants, vesting, and sales.
7. Add a likely batch warning when records have identical filing dates, similar values, or other observable batch characteristics.
8. Log successful checks, errors, and sent alerts.
9. Recover cleanly after temporary API or network errors.
10. Provide a test mode that uses a known historical record without sending duplicate live alerts.

Before writing the implementation, inspect the available Quiver connection and list the actual fields and methods you can use. Do not invent an endpoint, field, or streaming capability.

If the data source does not support event delivery, implement the most appropriate periodic check and explain the resulting delay.

Use configuration values for:
QUIVER credentials
TELEGRAM bot token
TELEGRAM destination chat
Check frequency
Watchlist
Minimum transaction value
Included insider titles
Cluster window

Alert specification:
[PASTE YOUR COMPLETED ALERT SPECIFICATION HERE]

After implementation, provide exact instructions to configure, test, start, stop, and monitor the service.

INSIDER PURCHASE ALERT

Ticker: [TICKER]
Company: [COMPANY]
Insider: [NAME]
Title: [TITLE]
Transaction date: [DATE]
Filing date: [DATE]
Shares: [COUNT]
Price: [PRICE]
Reported value: [VALUE]

Cluster: [COUNT] distinct buyers within [WINDOW]
Batch warning: [NONE OR EXPLANATION]
Record status: [NEW]

/prompts
  connection-test.md
  single-ticker-insiders.md
  purchase-cluster-screen.md
  batch-stress-test.md
  congressional-recent-filings.md
  political-exposure.md
  ownership-profile.md
  early-activity-screen.md
  alert-specification.md

I use Quiver data for [NEWS TRADING, FUNDAMENTAL RESEARCH, OR LONG TERM INVESTING].

My usual research universe is [MARKET, SECTOR, OR WATCHLIST].
My holding period is [PERIOD].
The risks I care about most are [RISKS].

Create 10 Quiver research prompts that fit this process. Each prompt must identify the required dataset, distinguish transaction dates from filing dates, request the underlying records, and include at least one alternative explanation for the observed pattern.