holistics/anfra-skills

anfra-lookup-values

Verify the exact stored values of a field before filtering on them, guarding against case sensitivity, abbreviations, and typos.

View source

How to use

  1. Copy the installation command and review the source package before installing it in your own agent environment.
  2. Check the source document for required tools, credentials and supported hosts.
  3. Describe your input and desired deliverable; review the output against the source material.
npx skills add https://github.com/holistics/anfra-skills --skill anfra-lookup-values

Before you start

  • Installation does not include a model subscription, API allowance or permission to access your files.
  • Repository stars and installation counts describe activity, not an independent quality or security test.
Original skill document

Rendered from the source repository. Headings, examples, code, tables, links, and referenced images are preserved.

Verifying field values before filtering

Before you filter on a value, confirm it actually exists in the data in that exact form. Filtering on a guessed literal ("Second-hand", "male", "VN") silently returns wrong or empty results when the stored form differs.

When to use

  • Any filter on a dimension whose exact stored values you haven't confirmed.
  • Especially categorical / text fields — and even "obvious" ones (a gender

column can still hold unexpected values like "f", "m", "F", "M").

How

  1. Identify the field(s) most likely to carry the concept (check field names,

labels, descriptions). Go from the most likely first.

  1. List the distinct values, narrowing with a case-insensitive

where(... ilike ...) filter. Batch 2–3 candidate patterns in a single query (joined with or) to go faster.

  1. Once you've identified the exact value(s), proceed to the real query. Don't

keep using lookups to compute the answer — write the AQL.

Command (swappable per scenario)

Default (Anfra) — unique() lists a dimension's distinct values; add a case-insensitive where(<model>.<field> ilike '%pattern%') to probe candidates, batching several patterns with or:

sh
# distinct values of a (small) column
anfra query --dataset <uname> 'unique(products.type)'

# batch 2–3 candidate patterns for "second-hand" in one query
anfra query --dataset <uname> "unique(products.type) | where(products.type ilike '%second%' or products.type ilike '%sh%' or products.type ilike '%used%')"
Replace this command for other runtimes if needed. The guidance above is stable; only the command changes.

If nothing matches

If you can't find the concept in any field, ask the user to clarify rather than guessing a value.


anfra skills 0.21.0

from this repository

More skills

All skills