Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Process JSON with jq and YAML/TOML with yq. Filter, transform, query structured data efficiently. Triggers on: parse JSON, extract from YAML, query config, Docker Compose, K8s manifests, GitHub Actions workflows, package.json, filter data.
.claude/skills/aiskillstore-data-processing/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -18% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 5% | 0% |
| case-21 | ✓→✓ | ＝ Same ✓ | 6% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | -10% | 0% |
Query, filter, and transform structured data (JSON, YAML, TOML) efficiently from the command line.
| Tool | Command | Use For | |------|---------|---------| | jq | jq '.key' file.json | JSON processing | | yq | yq '.key' file.yaml | YAML/TOML processing |
bash# Extract single field jq '.name' package.json # Extract nested field jq '.scripts.build' package.json # Extract from array jq '.dependencies[0]' package.json # Extract multiple fields jq '{name, version}' package.json # Navigate deeply nested jq '.data.users[0].profile.email' response.json # Filter by condition jq '.users[] | select(.active == true)' data.json # Transform each element jq '.users | map({id, name})' data.json # Count elements jq '.users | length' data.json # Raw string output jq -r '.name' package.json
bash# Extract field yq '.name' config.yaml # Extract nested yq '.services.web.image' docker-compose.yml # List all keys yq 'keys' config.yaml # List all service names (Docker Compose) yq '.services | keys' docker-compose.yml # Get container images (K8s) yq '.spec.template.spec.containers[].image' deployment.yaml # Update value (in-place) yq -i '.version = "2.0.0"' config.yaml # TOML to JSON yq -p toml -o json '.' config.toml
| Task | jq | yq | |------|----|----| | Get field | jq '.key' | yq '.key' | | Array element | jq '.[0]' | yq '.[0]' | | Filter array | jq '.[] \| select(.x)' | yq '.[] \| select(.x)' | | Transform | jq 'map(.x)' | yq 'map(.x)' | | Count | jq 'length' | yq 'length' | | Keys | jq 'keys' | yq 'keys' | | Pretty print | jq '.' | yq '.' | | Compact | jq -c | yq -o json -I0 | | Raw output | jq -r | yq -r | | In-place edit | - | yq -i |
For complete pattern libraries, load:
./references/jq-patterns.md - Arrays, filtering, transformation, aggregation, output formatting./references/yq-patterns.md - Docker Compose, K8s, GitHub Actions, TOML, YAML modification./references/config-files.md - package.json, tsconfig, eslint/prettier patterns| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-21 | pass→pass | 12,166 | 8,447 | -31% | 1 | 1 | 0% | 2,177 | 2,305 | +6% | 0 | 0 | — |
case-01 | fail→pass | 7,927 | 2,366 | -70% | 1 | 1 | 0% | 1,366 | 1,118 | -18% | 0 | 0 | — |
case-02 | pass→pass | 8,005 | 2,996 | -63% | 1 | 1 | 0% | 1,353 | 1,223 | -10% | 0 | 0 | — |
case-03 | pass→pass | 3,013 | 3,020 | +0% | 1 | 1 | 0% | 482 | 1,149 | +138% | 0 | 0 | — |
case-04 | pass→pass | 6,667 | 2,694 | -60% | 1 | 1 | 0% | 1,173 | 1,188 | +1% | 0 | 0 | — |
case-05 | pass→pass | 4,310 | 4,211 | -2% | 1 | 1 | 0% | 701 | 1,431 | +104% | 0 | 0 | — |
case-06 | pass→pass | 5,106 | 2,745 | -46% | 1 | 1 | 0% | 877 | 1,223 | +39% | 0 | 0 | — |
case-07 | pass→pass | 8,492 | 2,316 | -73% | 1 | 1 | 0% | 1,286 | 992 | -23% | 0 | 0 | — |
case-08 | pass→pass | 6,665 | 2,747 | -59% | 1 | 1 | 0% | 1,145 | 1,198 | +5% | 0 | 0 | — |
case-09 | pass→pass | 5,682 | 3,000 | -47% | 1 | 1 | 0% | 947 | 1,191 | +26% | 0 | 0 | — |
case-10 | pass→pass | 7,950 | 2,816 | -65% | 1 | 1 | 0% | 1,303 | 1,145 | -12% | 0 | 0 | — |
case-11 | fail→pass | 8,705 | 4,006 | -54% | 1 | 1 | 0% | 1,447 | 1,411 | -2% | 0 | 0 | — |
case-12 | pass→pass | 3,910 | 2,381 | -39% | 1 | 1 | 0% | 654 | 1,110 | +70% | 0 | 0 | — |
case-13 | pass→pass | 7,763 | 1,830 | -76% | 1 | 1 | 0% | 1,350 | 976 | -28% | 0 | 0 | — |
case-14 | fail→pass | 8,908 | 4,834 | -46% | 1 | 1 | 0% | 1,518 | 1,592 | +5% | 0 | 0 | — |
case-15 | pass→pass | 7,194 | 2,784 | -61% | 1 | 1 | 0% | 1,246 | 1,122 | -10% | 0 | 0 | — |
case-16 | pass→pass | 5,677 | 2,089 | -63% | 1 | 1 | 0% | 1,082 | 1,065 | -2% | 0 | 0 | — |
case-17 | pass→pass | 5,173 | 1,744 | -66% | 1 | 1 | 0% | 817 | 1,000 | +22% | 0 | 0 | — |
case-18 | pass→pass | 9,424 | 1,942 | -79% | 1 | 1 | 0% | 899 | 968 | +8% | 0 | 0 | — |
case-19 | pass→pass | 3,202 | 2,416 | -25% | 1 | 1 | 0% | 523 | 1,075 | +106% | 0 | 0 | — |
case-20 | pass→pass | 8,601 | 4,208 | -51% | 1 | 1 | 0% | 1,517 | 1,450 | -4% | 0 | 0 | — |
case-22 | pass→pass | 6,952 | 3,501 | -50% | 1 | 1 | 0% | 1,095 | 1,355 | +24% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +14 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.