structural-search
Search code by AST structure using ast-grep. Find semantic patterns like function calls, imports, class definitions instead of text patterns. Triggers on: find all calls to X, search for pattern, refactor usages, find where function is used, structural search, ast-grep, sg.
Install with Tessl CLI
npx tessl i github:NeverSight/skills_feed --skill structural-search100
Does it follow best practices?
Validation for skill structure
Evaluation results
100%
80%Legacy JavaScript Security Audit
Security vulnerability pattern detection
Uses ast-grep tool
0%
100%
eval pattern metavariable
0%
100%
innerHTML pattern metavariable
0%
100%
Loose equality pattern
0%
100%
JSON output flag
0%
100%
findings.json produced
100%
100%
Correct eval count
100%
100%
Correct innerHTML count
0%
100%
Count or list flag used
0%
100%
Reproducible commands
0%
100%
Without context: $0.4201 · 2m 8s · 17 turns · 22 in / 8,743 out tokens
With context: $0.7583 · 2m 49s · 36 turns · 39 in / 10,052 out tokens
90%
60%Modernise a Legacy Node.js Codebase
Large-scale refactoring with preview and apply
Uses ast-grep for migration
0%
100%
Preview flag used
0%
100%
Rewrite flag used
0%
0%
$NAME in var pattern
0%
100%
$_ in console.log pattern
0%
100%
Replacement value preserved
0%
100%
No var remains
100%
100%
No console.log remains
100%
100%
logger.info present
100%
100%
Preview before apply order
0%
100%
Without context: $0.2738 · 1m 1s · 19 turns · 25 in / 3,357 out tokens
With context: $0.6943 · 2m 11s · 36 turns · 38 in / 7,506 out tokens
100%
77%Automated Code Quality Rules for CI
Custom YAML lint rules with sg scan
sg scan command used
0%
100%
YAML rule files created
0%
100%
Rule has id field
0%
100%
Rule has language field
0%
100%
Rule has rule field
0%
100%
Rule has message field
0%
100%
eval rule pattern
0%
100%
innerHTML rule pattern
0%
100%
Scan results produced
100%
100%
Three violations detected
100%
100%
Without context: $0.4040 · 1m 41s · 26 turns · 28 in / 5,753 out tokens
With context: $0.5489 · 1m 45s · 33 turns · 286 in / 5,492 out tokens
- Evaluated
- Agent
- Claude Code
Table of Contents
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.