All Services

Returns editable Three.js WebGPU maze game source files for the vertical-27-d2 preset, configurable by seed, title, and accent color.

Returns editable Three.js WebGPU maze game source files for the vertical-27-d3 preset, configurable with a custom title, seed, and accent color.

Returns editable Three.js WebGPU/WebGL2 maze game source files using the vertical-27-d4 preset, with configurable title, seed, and accent color.

Returns real-time operational status and latency for major cross-chain bridges (Wormhole, Li.Fi, deBridge) with a confidence-weighted signal.

Returns 24-hour and 30-day fee or revenue totals for the three.ws AI agent marketplace, broken down by payment protocol.

Returns an AI-generated bullish/bearish/neutral trading signal with confidence score, rationale, headline, and market data for a given token

Launches a new token on Pump.fun (Solana) with optional vanity mint address, returning the mint address, transaction signature, and metadata URI

Runs a session lifecycle health check against the three.ws authentication system, verifying that session creation, expiry, refresh, and validation all pass with measured latencies.

Returns a list of cryptocurrency market exchanges with trust scores, 24h trading volume, and current BTC/USD price

Computes adjacent row-to-row changes for ticket and escalation count data across teams and dates, returning a local descriptive analytics report.

Computes a base-100 indexed series for ticket escalation or ticket volume data across teams and dates, enabling local descriptive analytics.

Computes category-frequency counts for ticket escalation records grouped by team and date, returning a bounded descriptive analytics report.

Computes a daily sum time series for ticket escalation and ticket volume data across teams, returning bounded descriptive analytics without forecasting or enrichment.

Generates a date-by-category pivot table for ticket escalation data, summarizing tickets or escalations across teams over time.

Fills missing numeric values (tickets or escalations) in up to 200 ticket-escalation records using zero, mean, or median imputation, returning a bounded local analytics report.

Computes a grouped descriptive statistics summary (count, mean, min, max, etc.) for ticket escalation data segmented by team and date, locally with no external enrichment.

Filters ticket escalation records to only those rows where a chosen numeric field (tickets or escalations) falls within a specified min–max range, returning up to 200 matched records.

Sorts up to 200 ticket escalation records by a chosen numeric field (tickets or escalations) and returns them in ranked order.

Audits up to 200 ticket escalation records for data quality issues, returning a bounded descriptive analytics report with no enrichment or forecasting.

Computes a secondary-measure weighted mean for up to 200 ticket escalation records, delivering a local descriptive analytics report with no enrichment or forecasting.

Computes confidence calibration bins for a ticket priority classifier using up to 200 labelled text records with predicted probabilities and actual labels.

Generates a fixed-holdout learning curve for a ticket priority classifier (low/normal/high) using up to 200 labelled text records and a naive-Bayes baseline.

Splits a labelled ticket-priority dataset (up to 200 records) into grouped train/test partitions for local ML evaluation and naive-Bayes baseline benchmarking.

Runs a grouped three-fold cross-validation on up to 200 labelled support ticket texts to evaluate a naive-Bayes classifier's accuracy at predicting low, normal, or high priority.

Audits labeled ticket priority predictions (low/normal/high) against actuals for up to 200 records, returning evaluation metrics and naive-Bayes baseline comparisons.

Computes the label distribution (class balance) across low, normal, and high priority labels in a supplied set of up to 200 labelled ticket text records.

Identifies low-confidence ticket priority predictions (low/normal/high) from a supplied batch of labelled records and returns a review queue of uncertain cases for human inspection.

Computes macro, micro, and weighted classification evaluation scores for a ticket priority (low/normal/high) classifier given up to 200 labelled text records with actual and predicted labels.

Computes the multiclass Brier score for a ticket priority classifier (low/normal/high) from up to 200 labelled records with predicted probabilities.

Computes multiclass log loss for a ticket priority classifier (low/normal/high) given up to 200 labelled records with predicted probabilities.

Evaluates classifier predictions for ticket priority (low/normal/high) across supplied labelled text records, broken down by group, using local naive-Bayes baselines and dataset tools.

Trains a local naive-Bayes text classifier on up to 200 labelled ticket records and classifies new query texts into low, normal, or high priority

Plays a move or manages a game of tic-tac-toe, returning the game state after each play.

Generates an AI-assisted schema-only CSV deduplication setup report and QA workbook in Markdown format, based on a text brief describing columns, delimiter, and dedupe key — no file uploads accepted.

Lists all TikTok accounts associated with a wallet, along with their current session health status.

Returns a reusable, substepped axis-separated wall-collision and movement ES module preconfigured with the balanced-23-d2 maze preset, with configurable title, seed, and accent color.

Returns a reusable, substepped, axis-separated wall-collision and movement ES module for the braided-11-d2 preset maze, with configurable title, seed, and accent color.

Returns a reusable ES module implementing substepped axis-separated wall-collision and movement logic for the braided-15-d3 preset maze, configurable by title, seed, and accent color.

Returns a reusable JavaScript ES module implementing substepped axis-separated wall-collision and movement logic with the preset open-11-d2 maze layout, configurable by title, seed, and accent color.

Returns a reusable, substepped axis-separated wall-collision and movement ES module pre-configured with the open-11-d4 maze preset, with configurable title, seed, and accent color.

Returns a reusable, substepped, axis-separated wall-collision and movement ES module for a preset maze game (open-15-d4) with configurable title, seed, and accent color.

Returns a reusable ES module implementing substepped, axis-separated wall-collision and movement logic for a preset maze layout (open-23-d3), with configurable title, seed, and accent color.

Returns a reusable, substepped, axis-separated wall-collision and movement ES module for the preset vertical-11-d3 maze game, with configurable title, seed, and accent color.

Returns a reusable ES module implementing substepped, axis-separated wall-collision and movement logic for a vertical-11-d4 preset maze, with configurable title, seed, and accent color.

Returns a reusable, substepped, axis-separated wall-collision and movement ES module for the preset vertical maze layout 'vertical-23-d2', with configurable title, seed, and accent color.

Converts a batch of time-entry records (hours or billable_hours) into a base-100 indexed series anchored to a chosen row, enabling relative comparison across dates and projects.

Computes category frequency counts for time entry records grouped by project, returning how often each category appears across up to 200 supplied rows.

Computes a daily sum time series for time entry records, aggregating hours or billable_hours by date across up to 200 rows.

Generates a date-by-category pivot table for time entry records, aggregating hours or billable hours by project across dates.

Identifies duplicate row_id keys across a batch of time entry records and returns a data-quality report flagging duplicates.

Computes an equal-width histogram over a numeric field (hours or billable_hours) from up to 200 time entry records, returning bin counts and boundaries.

Removes exact duplicate rows from a batch of time entry records based on all fields matching identically, returning only unique records.

Computes a grouped numeric summary (e.g. sum, mean) of hours or billable_hours fields across time entry records, grouped by project.

Compares two keyed datasets of time entry records (up to 200 rows each) and returns data-quality and numeric difference analysis between them.

Filters time entry records (hours or billable_hours) to only those rows whose numeric value falls within a specified minimum–maximum range.

Computes a trailing (rolling) mean series over time entry records for hours or billable_hours fields, using a configurable window size.

Computes the two-measure (Pearson-style) correlation between hours and billable_hours across up to 200 time entry records, locally, with no external enrichment.

Extracts all distinct values of a specified HTML attribute from <time> elements on a given webpage or supplied HTML snippet.

Extracts <time> HTML elements from a web page or supplied HTML that contain a specified literal term match, with optional attribute filtering.

Extracts text content from HTML <time> elements matching a given attribute filter and compares it against a supplied set of previous text values to identify changes.

Analyzes time series completeness records to produce two reports: adjacent row change analysis and a date-by-category pivot table, for up to 200 rows.

Analyzes time series completeness records (observed vs expected points) by computing adjacent row changes and an equal-width histogram in a single API call.

Analyzes time series completeness data by computing adjacent row differences and filling missing values (zero, mean, or median) for observed_points or expected_points fields across up to 200 rows.

Analyzes time series completeness records by computing adjacent row changes and two-measure correlation between observed_points and expected_points for up to 200 rows.

Computes a Base-100 indexed time series and a date-by-category pivot table from supplied time series completeness records (observed vs expected data points), returning two analytical reports in one call.

Generates a Base-100 index series and an equal-width histogram for time series completeness data (observed vs expected data points) supplied by the caller.

Computes a base-100 indexed series and fills missing values (zero, mean, or median) for time series completeness data containing observed vs. expected point counts.

Computes a Base-100 index series and a secondary-measure weighted mean for time series completeness records (observed vs expected data points), returning two analytical reports for up to 200 rows.

Computes a base-100 indexed series and a two-measure correlation between observed_points and expected_points for up to 200 time series completeness records.

Analyzes time series completeness data by computing category frequencies and adjacent row-to-row changes for up to 200 rows, returning two local analytical reports per call.

Analyzes time series completeness records to produce category frequency distributions and a base-100 indexed series report for up to 200 rows.

Analyzes time series completeness records by computing category frequencies and a daily sum series for observed vs expected data points across up to 200 rows.

Analyzes time series completeness records by computing category frequencies and a date-by-category pivot table for observed vs. expected data points.

Generates two analytical reports — category frequency counts and an equal-width histogram — for time series completeness data (observed vs. expected points) across up to 200 rows.

Runs two local analytical reports on time series completeness data: category frequency distribution by group and keyed row-by-row comparison between two datasets (up to 200 rows each).

Analyzes time series completeness records by computing category frequencies for dataset groups and numeric row ordering for observed/expected point values, delivering two analytical reports for up to 200 rows.

Computes category-level frequencies and a secondary-measure weighted mean for up to 200 time series completeness records (observed vs. expected points), returning two independent analytical reports.

Analyzes time series completeness records by computing category frequencies and a trailing mean series across up to 200 rows in two independent analytical reports.

Analyzes time series completeness records (observed vs expected data points) to produce two reports: category frequency distribution by dataset group and correlation analysis between observed_points and expected_points.

Analyzes time series completeness records by producing a daily sum series report and a date-by-category pivot table for observed vs expected data points.

Analyzes time series completeness records by producing a daily sum series and an equal-width histogram for observed_points or expected_points, using local analytical engines on up to 200 rows.

Analyzes daily time series completeness records and fills missing values using zero, mean, or median imputation, producing two analytical reports for up to 200 rows.

Analyzes time series completeness records by computing a daily sum series and comparing two keyed datasets across up to 200 rows, returning two independent analytical reports.

Computes daily sum and secondary-measure weighted mean analytics for time series completeness records (observed vs expected data points) across up to 200 rows in two independent reports.

Computes a daily sum series and a two-measure correlation report for time series completeness data (observed vs expected points) over up to 200 rows.

Pivots time series completeness records (observed vs expected data points) by date and category, then fills missing values using zero, mean, or median imputation — delivering two analytical reports for up to 200 rows.

Analyzes time series completeness records to produce a duplicate key report and a base-100 indexed series for up to 200 rows.

Analyzes up to 200 time series completeness records to produce a duplicate key report and category frequency breakdown for observed vs. expected data points.

Analyzes supplied time series completeness records to produce a duplicate key report and a daily sum series for observed or expected data points, for up to 200 rows.

Analyzes time series completeness records to produce a duplicate key report and a date-by-category pivot table for up to 200 rows.

Analyzes time series completeness records (observed_points, expected_points) to produce a duplicate key report and an equal-width histogram for up to 200 rows.

Analyzes up to 200 time series completeness records to produce a duplicate key report and a grouped numeric summary of observed vs. expected data points.

Analyzes up to 200 time series completeness records to produce a duplicate key report and IQR-based outlier flags for observed_points or expected_points values.

Analyzes time series completeness records (observed_points, expected_points) to produce a duplicate key report and numeric distribution for up to 200 rows.

Analyzes time series completeness records to produce a duplicate key report and a numeric interval filter report for observed/expected data points across up to 200 rows.

Analyzes up to 200 rows of time series completeness records to produce a duplicate key report and a two-measure correlation analysis between observed_points and expected_points.

Analyzes time series completeness records (observed vs expected data points) by building an equal-width histogram and filling missing values, producing two analytical reports for up to 200 rows.

Generates an equal-width histogram and IQR-based outlier flags for time series completeness data (observed vs expected points) across up to 200 rows, producing 2 independent analytical reports per call.

Runs two local statistical reports (equal-width histogram and two-measure correlation) on time series completeness data (observed vs. expected points) for up to 200 rows.

Analyzes up to 200 time series completeness records to produce two reports: exact duplicate row detection and adjacent-row change detection for observed_points or expected_points.