timesfm-forecasting

Zero-shot time series forecasting with Google's TimesFM foundation model. Use this skill when forecasting ANY univariate time series — sales, sensor readings, stock prices, energy demand, patient vitals, weather, or scientific measurements — without training a custom model. Supports both basic forecasting and advanced covariate forecasting (XReg) with dynamic and static exogenous variables. Automatically checks system RAM/GPU before loading the model, validates dataset fit before processing, supports CSV/DataFrame/array inputs, and returns point forecasts with calibrated prediction intervals. Includes a preflight system checker script that MUST be run before first use to verify the machine can load the model and handle your specific dataset.

Quality

91%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

82%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

Highly actionable content with strong workflows and validation checkpoints, well-organized into real one-level reference files. The main weakness is conciseness: the system-checker and hardware-requirement material is repeated across multiple sections and could be consolidated into the referenced system_requirements.md.

Suggestions

Consolidate the three separate system-checker sections (mandatory preflight, installation Step 1, Available Scripts) into one and defer detail to references/system_requirements.md to reduce repetition.

Move the duplicate RAM/GPU/disk check bullet list and the full ForecastConfig parameter table into references/api_reference.md, keeping only the must-set flags inline.

Trim the mermaid flowchart and hardware-requirements table to a single concise summary since both restate the preflight script's own checks.

Dimension	Reasoning	Score
Conciseness	Mostly efficient with executable code, but repeated material — the system-checker section appears three times, RAM/GPU/disk checks listed twice, and verbose hardware tables and mermaid diagrams restate things Claude can infer — adds padding that could be tightened.	3 / 5
Actionability	Copy-paste-ready executable code spans the common cases (minimal forecast, covariates, batch, accuracy eval, anomaly detection) plus concrete CLI commands and exact parameter tables with 'when to change' guidance.	5 / 5
Workflow Clarity	The mandatory preflight checker with its decision flowchart, dataset-fit estimation, an explicit Quality Checklist, and a Validation & Verification section with passing assertions provide explicit checkpoints and feedback loops for the batch/destructive operations.	5 / 5
Progressive Disclosure	Good one-level-deep structure: SKILL.md overviews and signals references/api_reference.md, data_preparation.md, and system_requirements.md, all of which exist as real files; minor gaps from some detail (full ForecastConfig reference, anomaly tables) being inlined rather than moved out.	4 / 5
	Total	17 / 20 Passed

Description

100%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, specific description that clearly states both what the skill does and when to use it, with concrete trigger terms covering many natural forecasting scenarios. Third-person voice is maintained throughout.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'zero-shot time series forecasting', 'covariate forecasting (XReg) with dynamic and static exogenous variables', 'returns point forecasts with calibrated prediction intervals' — covering the full capability surface comprehensively.	5 / 5
Completeness	Explicit 'Use this skill when forecasting ANY univariate time series' trigger plus a clear 'what' (zero-shot forecasting, covariate support, calibrated intervals, system preflight) jointly answer both what and when with concrete phrases.	5 / 5
Trigger Term Quality	Names concrete user-facing targets — 'sales, sensor readings, stock prices, energy demand, patient vitals, weather, or scientific measurements' — alongside natural terms like 'forecasting' and 'prediction intervals', giving comprehensive keyword coverage.	5 / 5
Distinctiveness Conflict Risk	The TimesFM foundation-model forecasting niche with explicit covariate and preflight-checker triggers is highly distinct and unlikely to fire for unrelated skills.	5 / 5
	Total	20 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (512 lines); consider splitting into references/ and linking	Warning

	Total	15 / 16 Passed

Repository: google-research/timesfm
Path: timesfm-forecasting/SKILL.md
Commit: 8cb0628

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.