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. Automatically checks system RAM/GPU before loading the model, 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. For classical statistical time series models (ARIMA, SARIMAX, VAR) use statsmodels; for time series classification/clustering use aeon.

Quality

93%

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

86%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.

A thorough, highly actionable skill body with excellent executable examples, gated preflight, and strong validation/checklist feedback loops. Its main weakness is verbosity — duplicated config and quantile-index material and a stray repeated code block — plus references to examples/ directories that do not exist in the bundle.

Suggestions

Remove the duplicate content: delete the second forecast_with_covariates() code block after the anomaly-detection table (it repeats the covariates example), and consolidate the ForecastConfig parameter list and its table into one place, deferring the full table to references/api_reference.md.

Tighten the quantile-index material: the index-to-quantile mapping is explained in 'Understanding the Output', restated in the Common Mistakes section, and implied in the Quality Checklist — keep one canonical table and reference it.

Verify the examples/ directories referenced in the body (global-temperature, anomaly-detection, covariates-forecasting) actually exist in the bundle, or replace those pointers with references to the real bundle files to avoid dead navigation.

Dimension	Reasoning	Score
Conciseness	The body is ~775 lines with noticeable padding: the ForecastConfig parameters are fully enumerated inline AND again as a table, the quantile index table appears twice (Understanding the Output + Common Mistakes), and a stray duplicate forecast_with_covariates() block repeats after the anomaly-detection section — efficient in spots but clearly could be tightened.	3 / 5
Actionability	Copy-paste-ready executable code throughout — the minimal 5-line example, CSV forecast, covariates, batch, evaluation, memory-tuning — plus concrete CLI invocations for scripts/check_system.py and forecast_csv.py covering the common cases.	5 / 5
Workflow Clarity	Mandatory preflight is gated behind an explicit CRITICAL warning and a Mermaid decision flowchart; the single-series workflow is numbered 1–7 with validation (system check before load, compile before forecast), and the Validation & Verification section plus Quality Checklist provide explicit checkpoints and feedback loops for batch/destructive operations.	5 / 5
Progressive Disclosure	Good one-level-deep structure: SKILL.md is an overview with a clear Reference Documentation table pointing to references/system_requirements.md, api_reference.md, and data_preparation.md (all real files), plus listed scripts; however some bulk reference material (full ForecastConfig + parameter table, quantile anatomy) is inlined in SKILL.md rather than deferred, and the body references examples/ directories that are not present in the bundle.	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 highly specific, third-person description that crisply states what the skill does, when to use it, and when NOT to use it via explicit de-routing to statsmodels and aeon. It hits concrete actions, natural trigger terms, and a distinct niche with minimal conflict risk.

Dimension	Reasoning	Score
Specificity	Names multiple concrete actions — 'checks system RAM/GPU before loading the model', 'supports CSV/DataFrame/array inputs', 'returns point forecasts with calibrated prediction intervals', 'Includes a preflight system checker script' — covering the capability surface comprehensively.	5 / 5
Completeness	Explicitly answers both 'what' (zero-shot forecasting returning point forecasts with calibrated intervals) and 'when' via the direct 'Use this skill when forecasting ANY univariate time series ...' clause, with concrete trigger phrases and an explicit negative boundary.	5 / 5
Trigger Term Quality	Rich natural vocabulary users would say — 'sales, sensor readings, stock prices, energy demand, patient vitals, weather', 'prediction intervals', 'zero-shot forecasting' — plus concrete format terms CSV/DataFrame/array; matches the comprehensive-synonym anchor.	5 / 5
Distinctiveness Conflict Risk	Clear niche (TimesFM foundation-model zero-shot forecasting) with explicit de-routing to statsmodels for ARIMA/SARIMAX/VAR and aeon for classification/clustering, minimizing overlap with adjacent skills.	5 / 5
	Total	20 / 20 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (797 lines); consider splitting into references/ and linking	Warning
metadata_version	'metadata.version' is missing	Warning

	Total	14 / 16 Passed

Repository: foryourhealth111-pixel/Vibe-Skills
Path: bundled/skills/timesfm-forecasting/SKILL.md
Commit: ddcaa2a

Reviewed: 7 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.