Goodeye Labs
Use Goodeye
Public Registry

Reusable workflows for outcome-aligned agents

Search public Goodeye templates, copy a CLI command, and let your agent fetch the runbook through Goodeye.

Showing results for “data”.Clear
@randalolson/text-to-sql-query-builderv1

text-to-sql-query-builder

Use when the user asks an agent (in a code or data repo) to translate a natural-language question into dialect-correct SQL with a machine-checkable rationale. The workflow has the agent discover schema and dialect from the repo before asking the user, drafts the query plus a structured rationale, then runs five programmatic verifiers (parse, explain, column-match, rationale-coverage, write-guard). Use when an agent must answer questions like "top 10 customers by revenue last 30 days" against a known schema.

Outcome

Produce a SQL query and a structured rationale that pass parse, explain, column-match, rationale-coverage, and write-guard verifiers, after the agent discovers schema and dialect from repo context.

sqltext-to-sqldatacode-generationverifierssqlglotduckdb
View usage snippets
CLI
goodeye templates get @randalolson/text-to-sql-query-builder
@randalolson/high-signal-chart-workflowv1

High-Signal Chart Workflow

Idea -> dataset discovery -> three parallel chart-type variants -> winner selection -> verifier loop: a local programmatic chart check, then Truesight's chart-design semantic evaluator. Artifacts land in a local run directory.

Outcome

Produce a publication-quality, high-signal chart from a one-line idea, passing both programmatic layout checks and the remote design evaluator.

data-visualizationchart-design-evaltruesightdemo
View usage snippets
CLI
goodeye templates get @randalolson/high-signal-chart-workflow