Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    market data integrity validator

    1

    Validate market data integrity by detecting sequence gaps, stale timestamps, and crossed quotes in feed exports.

    $15

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    market data integrity validator

    market data integrity validator

    Example session with this skill installed

    Validate the attached quote stream for BTC-USDT and ETH-USDT from the Binance export. The expected as-of time is 2024-05-20T10:00:00Z. Check for gaps and stale quotes.

    • Read your context and instructions
    • Compiled the market data integrity

    The market data integrity check is complete. I have analyzed the quote stream for two symbols, identifying significant sequence gaps in the BTC-USDT feed and three stale quotes in the ETH-USDT feed.

    Market Data Validation Report

    MetricBTC-USDT (Binance)ETH-USDT (Binance)
    Total Messages15,40212,890
    Chronological Gaps4 (Totaling 122ms)0
    Sequence Violations12 (Dropped packets)0
    Stale Timestamps03 (>500ms old)
    Crossed Quotes1 (Bid > Ask at 10:00:04)0
    Integrity Score92.4%99.8%

    Detailed Findings

    BTC-USDT Sequence Jump: A major sequence break occurred between sequence IDs 88294 and 88307. This indicates a burst of packet loss at the ingest layer.

    ETH-USDT Latency: Three quotes showed a server_t vs receipt_t delta exceeding the 500ms threshold. These occurred during a period of high volatility at 09:58:12Z.

    Crossed Quote: A single instance where Bid (67050.5) exceeded Ask (67049.0) was detected in the BTC feed; likely a venue matching engine artifact or malformed export line.

    {
      "report_version": "1.0.2",
      "as_of": "2024-05-20T10:00:00Z",
      "summary": {
        "status": "WARNING",
        "critical_issues": 2
      },
      "recommendations": [
        "Resubscribe to BTC-USDT to reset sequence counters.",
        "Investigate local network jitter for ETH-USDT stale receipt times."
      ]
    }
    

    Next steps

    1. Review the specific sequence IDs in the report.json to correlate with network logs.
    2. Filter the BTC-USDT dataset to remove the crossed quote entry before starting backtesting.
    3. Check feed handler logs for the 10:00:04Z timestamp.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    The problem

    Corrupted market data feeds lead to bad fills and broken backtests. Identifying chronological gaps, stale timestamps, or crossed quotes manually in high-volume JSON or CSV exports is nearly impossible.

    What it does

    • Detects stale, future, and out-of-sequence timestamps against a defined as-of time.
    • Identifies sequence number gaps and duplicates while preserving original receipt order.
    • Validates bid/ask spreads to flag crossed quotes or anomalous pricing data.
    • Audits feed inventory by comparing observed streams against expected venue symbols.
    • Provides resubscription recommendations based on detected data integrity failures.

    Frameworks & tools

    Python 3 for analysis scripts, JSON/CSV for data input and reporting.

    Why this beats prompting it yourself

    General LLMs often try to "fix" data by sorting it, which hides the very sequence errors you need to find. This skill enforces strict receipt-order preservation and applies venue-specific sequence logic that basic prompts ignore.

    Use cases

    • Post-mortem analysis of trading slippage caused by data lag.
    • Pre-processing historical market data before running backtests.
    • Validating the integrity of a new third-party data provider's API export.

    Known limitations

    Handles top-of-book quotes only. Does not include live WebSocket connectivity, exchange snapshot resync, or automated trading kill-switch logic.

    Supports top-of-book quotes only. No live execution or WebSocket clients. Requires manual mapping of external data to the supported input contract.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 1 month ago

    Requires

    Python

    Permissions

    Terminal / Shell
    Listed1 month ago

    What's inside

    Frequently Asked Questions

    It checks supplied normalized quote records for stale or future timestamps, crossed bid/ask quotes, out-of-order timestamps, and optional sequence gaps or replays. It does not ingest CSV directly or guarantee that downstream backtests are safe.

    The helper runs with Python 3.10+ and the documented normalized JSON contract. An agent or platform may prepare that input, but no CSV importer, feed connector, or platform-specific integration is included.

    Unlike general prompts that may inadvertently reorder data, this skill preserves the original receipt order to identify true sequence failures and applies specific bid/ask validation logic.

    The package includes the instructions, local Python validator, normalized JSON contract, synthetic valid and adverse fixtures, example report, runtime notes, smoke-test script, source notes, and logo.

    No, this skill is designed for post-trade analysis and historical data validation; it does not support live WebSocket streams or real-time trading execution.

    Yes. Convert the data to the documented JSON contract with an explicit as-of timestamp, age limit, sequence policy, unique record IDs, venue, symbol, timestamp, positive bid and ask, and sequence numbers when contiguous-sequence checking is enabled. CSV exports must be mapped externally.

    New in Trading & Investing