timesfm-forecasting

Zero-shot time series forecasting with Google's TimesFM foundation model. Use for any univariate time series (sales, sensors, energy, vitals, weather) without training a custom model. Supports CSV/DataFrame/array inputs with point forecasts and prediction intervals. Includes a preflight system checker script to verify RAM/GPU before first use.

Quality

83%

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

75%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 runnable examples, a strong preflight validation workflow, and well-signaled reference files. The main weaknesses are a duplicated covariates code block and two referenced example directories that are absent from the bundle.

Suggestions

Remove the duplicated covariates code block under the Anomaly Detection section (lines ~260–269) — it is identical to the Covariates example and is misplaced there.

Create the referenced examples/anomaly-detection/ and examples/covariates-forecasting/ directories, or remove/adjust the 'See examples/...' pointers so no referenced path dangles.

Consolidate the repeated system-check guidance (overview, preflight section, installation Step 1, and Available Scripts all restate it) into one authoritative preflight section to tighten the body.

Dimension	Reasoning	Score
Conciseness	Mostly efficient with useful code and tables, but the covariates code block is duplicated verbatim (once under Covariates, once mis-placed under Anomaly Detection) and the system check is restated in four places (overview, preflight section, installation Step 1, Available Scripts), so it could be tightened.	3 / 5
Actionability	Fully copy-paste-ready guidance throughout — a 5-line runnable minimal example, CSV/covariate/anomaly code blocks with documented output shapes, and concrete CLI commands with specific args ('python scripts/forecast_csv.py input.csv --horizon 24 --date-col date --value-cols sales,revenue --output forecasts.csv').	5 / 5
Workflow Clarity	Clear sequenced install workflow (Steps 1–4) plus a preflight decision flowchart with explicit BLOCKED/READY validation states and remediation, but result/output validation (e.g. NaN checks) is referenced in pitfalls rather than wired into the main forecast workflow as a checkpoint.	4 / 5
Progressive Disclosure	Good one-level-deep structure: the body links to 7 reference files and 2 scripts, all of which resolve, each with a one-line description. Held below 5 because the body also references examples/anomaly-detection/ and examples/covariates-forecasting/ directories that do not exist in the bundle.	4 / 5
	Total	16 / 20 Passed

Description

92%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 precise, well-constructed description that explicitly states capabilities and triggers in third person with concrete domain examples. It cleanly answers both what the skill does and when to use it.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'Zero-shot time series forecasting', 'Supports CSV/DataFrame/array inputs with point forecasts and prediction intervals', and 'preflight system checker script to verify RAM/GPU' — giving comprehensive capability coverage.	5 / 5
Completeness	Explicitly answers 'what' (zero-shot forecasting with TimesFM, input formats, intervals, system checker) and 'when' via the concrete 'Use for any univariate time series (sales, sensors, energy, vitals, weather) without training a custom model' trigger.	5 / 5
Trigger Term Quality	Strong natural domain terms ('sales, sensors, energy, vitals, weather', 'prediction intervals', 'zero-shot forecasting') that users would say, but no file extensions (e.g. .csv) and a few common synonyms (demand, forecast) are missing, so it sits just below comprehensive.	4 / 5
Distinctiveness Conflict Risk	A clear niche — zero-shot univariate foundation-model forecasting with TimesFM — with triggers specific enough to avoid overlap with statsmodels/scikit-learn style skills.	5 / 5
	Total	19 / 20 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: K-Dense-AI/scientific-agent-skills
Path: skills/timesfm-forecasting/SKILL.md
Commit: b2a92ba

Reviewed: 3 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.