agentsclimarketplace
Sign in

Synchronized multi language test data directory

Skill kjuhwa/skills-hub/skills/testing/synchronized-multi-language-test-data-directory

Single tests_data/ tree with one subdir per content type, shared across every language test suite, plus reference outputs and known-failure buckets.From its SKILL.md

Install
npx -y skills add kjuhwa/skills-hub --skill synchronized-multi-language-test-data-directory

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

1.8 KB, 299 tokens by cl100k_base, as published. Nobody here has run it

Synchronized Multi Language Test Data Directory

Trigger: Multi-language project where every binding must be tested against the same corpus to guarantee parity.

Steps

  • Organize tests_data/basic/<content-type>/ with 1–3 small sample files per type.
  • Add tests_data/current_missdetections/ for known-bad cases and tests_data/previous_missdetections/ for regressions.
  • Commit reference outputs (predictions + features) under tests_data/reference/ as JSONL.
  • In every binding's test suite, iterate over the subdirs and verify detection.
  • Set .gitattributes to keep binary samples binary and text samples LF-only — CRLF auto-conversion will corrupt your corpus.
  • Periodically audit current_missdetections/ and graduate passing files into basic/.

Counter / Caveats

  • Test files inflate repo size; cap at ~100KB each or use git-lfs.
  • Some types need rich examples (PDF/DOCX with text+images+metadata); others need a single byte.
  • CRLF/LF line ending drift silently breaks reference comparisons.
  • Maintaining the corpus is manual labor; budget time per release to keep it healthy.

Source

Extracted from magika (https://github.com/google/magika.git @ main).

Files of interest:

  • tests_data/basic/ (per-content-type subdirs)
  • tests_data/features_extraction/
  • python/tests/
  • go/cli/cli_test.go

What ships with it: 1 file

923 B alongside SKILL.md

Keep looking

Skills are one crate of 326,750. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.