- Home
- Skills
- Trading & Investing
- alternative data event corroborator
Works with the AI Agents you already use
alternative data event corroborator
Validate and score alternative data anomalies using independent source corroboration and z-score analysis.
$13
alternative data event corroborator
Example session with this skill installed
Analyze the anomaly in the provided retail_footfall.json against credit_card_velocity.json. Identify if the 15% spike is corroborated by independent source groups.
- Read your context and instructions
- Compiled the alternative data event
I have completed the corroboration analysis for the retail footfall anomaly. The analysis indicates a high z-score with corroboration across two independent source groups, though significant reporting latency was observed.
Alternative-Data Corroboration Report
| Metric | Value | Independence Group | Notes |
|---|---|---|---|
| Target Event | 15.2% Footfall Spike | Group A (Sensor) | Observed at T+0, reported at T+2h |
| Corroborating Data | 12.8% Transaction Lift | Group B (Payment Processor) | Observed at T+0, reported at T+24h |
| Baseline Mean | 2.1% Growth | N/A | 30-day trailing window |
| Z-Score | 4.82 | N/A | High statistical significance |
| Source Latency | 22 Hours | N/A | Credit card data lag vs sensors |
| Independence Status | Verified | 2 Groups | No shared underlying syndication detected |
Interpretation
The footfall anomaly is corroborated by transaction velocity. However, the transaction data arrived with a 22-hour delay relative to the sensor data. The z-score of 4.82 suggests the move is non-random, but the operator should account for selection bias in the sensor placement before scaling.
Next steps
- Verify if the seasonal basis for the 30-day window accounts for the recent holiday weekend.
- Map the
store_identities in the transaction dataset to confirm 1:1 coverage with sensor locations. - Run the primary-source follow-up script to check for regional power outages that might impact sensor uptime.
Connects securely to your tools. The creator never sees your data.
About this skill
The problem
Traders often struggle to verify if a numeric anomaly in alternative data is a genuine signal or just noise from a single, unreliable source. Relying on uncorroborated data leads to false positives and poor execution decisions.
What it does
- Normalizes units, entity mapping, and seasonal basis across disparate datasets.
- Assigns independence groups to sources to prevent double-counting syndicated reports.
- Calculates z-scores for numeric anomalies to measure statistical deviation from the baseline.
- Tracks reporting latency by recording the gap between event occurrence and data availability.
- Generates ranked event candidates with specific coverage and selection bias caveats.
Frameworks & tools
Python 3 for analysis scripts, JSON for data contracts, and standard statistical methods for z-score calculations.
Why this beats prompting it yourself
Standard LLM prompts often hallucinate correlations or fail to account for source independence. This skill enforces a rigorous methodology that separates operator assertions from helper calculations, ensuring you don't treat two copies of the same report as independent confirmation.
Use cases
- Verifying supply chain disruptions by cross-referencing shipping logs with inventory data.
- Analyzing foot traffic anomalies against point-of-sale datasets.
- Corroborating private company revenue estimates using alternative payroll and tax data.
Known limitations
Supports numeric series only. Does not perform satellite/web ingestion, automatic entity resolution, or causal inference. Requires pre-mapped data following the internal contract.
Limited to numeric series analysis. No built-in satellite/web scraping or automatic entity resolution. Requires manual mapping to the runtime contract.
How to use
- 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
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 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