oxc-walker-skilld
ALWAYS use when writing code importing \"oxc-walker\". Consult for debugging, best practices, or modifying oxc-walker, oxc walker.
Works with
---
name: oxc-walker-skilld
description: ALWAYS use when writing code importing \"oxc-walker\". Consult for debugging, best practices, or modifying oxc-walker, oxc walker.
license: MIT
---
# oxc-project/oxc-walker `oxc-walker`
**Version:** 0.7.0 (Jan 2026)
**Deps:** magic-regexp@^0.10.0
**Tags:** latest: 0.7.0 (Jan 2026)
**References:** [package.json](./.skilld/pkg/package.json) — exports, entry points • [README](./.skilld/pkg/README.md) — setup, basic usage • [GitHub Issues](./.skilld/issues/_INDEX.md) — bugs, workarounds, edge cases • [Releases](./.skilld/releases/_INDEX.md) — changelog, breaking changes, new APIs
## Search
Use `skilld search` instead of grepping `.skilld/` directories — hybrid semantic + keyword search across all indexed docs, issues, and releases. If `skilld` is unavailable, use `npx -y skilld search`.
```bash
skilld search "query" -p oxc-walker
skilld search "issues:error handling" -p oxc-walker
skilld search "releases:deprecated" -p oxc-walker
```
Filters: `docs:`, `issues:`, `releases:` prefix narrows by source type.
## API Changes
This section documents version-specific API changes -- prioritize recent major/minor releases.
- NEW: `ScopeTrackerVariable` -- exported class in v0.7.0, enables `instanceof` checks on results from `scopeTracker.getDeclaration()`. Previously internal-only; LLMs will not know these classes are importable [source](./.skilld/releases/v0.7.0.md#features)
- NEW: `ScopeTrackerFunction` -- exported class in v0.7.0, represents function declarations tracked by `ScopeTracker`. Import directly from `oxc-walker` for type narrowing [source](./.skilld/releases/v0.7.0.md#features)
- NEW: `ScopeTrackerImport` -- exported class in v0.7.0, represents import declarations tracked by `ScopeTracker`. Has `.importNode` property for accessing the full `ImportDeclaration` [source](./.skilld/releases/v0.7.0.md#features)
- NEW: `ScopeTrackerFunctionParam` -- exported class in v0.7.0, represents function parameters tracked by `ScopeTracker`. Has `.fnNode` for accessing the parent function node [source](./.skilld/releases/v0.7.0.md#features)
- NEW: `ScopeTrackerIdentifier` -- exported class in v0.7.0, represents plain identifiers (e.g. class/enum names) tracked by `ScopeTracker` [source](./.skilld/releases/v0.7.0.md#features)
- NEW: `ScopeTrackerCatchParam` -- exported class in v0.7.0, represents catch clause parameters tracked by `ScopeTracker`. Has `.catchNode` for the parent `CatchClause` [source](./.skilld/releases/v0.7.0.md#features)
- NEW: `getUndeclaredIdentifiersInFunction(node)` -- exported in v0.7.0, returns `string[]` of identifier names referenced but not declared within a `Function` or `ArrowFunctionExpression` node [source](./.skilld/releases/v0.7.0.md#features)
- NEW: `isBindingIdentifier(node, parent)` -- exported in v0.7.0, returns `boolean` indicating whether a node is a binding identifier given its parent context [source](./.skilld/releases/v0.7.0.md#features)
- DEPRECATED: `ScopeTrackerFunctionParam.start` -- use `.fnNode.start` instead. The representation of this position may change in future versions [source](./.skilld/pkg/dist/index.d.ts:L279)
- DEPRECATED: `ScopeTrackerFunctionParam.end` -- use `.fnNode.end` instead. The representation of this position may change in future versions [source](./.skilld/pkg/dist/index.d.ts:L283)
- NEW: `ScopeTrackerNode` type -- union type exported in v0.7.0: `ScopeTrackerFunctionParam | ScopeTrackerFunction | ScopeTrackerVariable | ScopeTrackerIdentifier | ScopeTrackerImport | ScopeTrackerCatchParam`. Use for typing `getDeclaration()` results [source](./.skilld/releases/v0.7.0.md#features)
- BREAKING: `oxc-parser` peer dependency -- v0.6.0 requires `oxc-parser` v0.98.0+. Older oxc-parser versions are no longer compatible [source](./.skilld/releases/v0.6.0.md#features)
**Also changed:** `WalkOptions` type exported v0.7.0 · `WalkerEnter` type exported v0.7.0 · `WalkerLeave` type exported v0.7.0 · `WalkerCallbackContext` type exported v0.7.0 · `WalkerThisContextEnter` type exported v0.7.0 · `WalkerThisContextLeave` type exported v0.7.0 · `Identifier` type exported v0.7.0 · `ScopeTrackerOptions` type exported v0.7.0
## Best Practices
- Use regular function expressions (not arrow functions) for `enter`/`leave` callbacks -- `this.skip()`, `this.replace()`, and `this.remove()` are bound via `this` context and won't work with arrow functions [source](./.skilld/pkg/dist/index.d.ts:L88:90)
```ts
// correct
walk(ast, { enter(node) { this.skip() } })
// broken -- this.skip is undefined
walk(ast, { enter: (node) => { this.skip() } })
```
- Use `parseAndWalk` over separate `parseSync` + `walk` when you only need a single pass -- it returns `ParseResult` (including `program`) so you can still re-walk the AST if needed later [source](./.skilld/pkg/README.md#parse-and-walk-directly)
- For hoisted declaration resolution, use the two-pass pattern: create `ScopeTracker({ preserveExitedScopes: true })`, run `parseAndWalk` as a pre-pass to collect all declarations, call `scopeTracker.freeze()`, then `walk(program, { scopeTracker })` for analysis -- without this, `getDeclaration()` returns `null` for identifiers declared after their usage in source order [source](./.skilld/pkg/README.md:L170:203)
- Do not use `this.replace()` or `this.remove()` when a `ScopeTracker` is attached -- the scope tracker will not update its internal declarations, leaving stale scope data [source](./.skilld/pkg/README.md#thisreplacenewnode)
- `this.remove()` takes precedence over `this.replace()` -- if both are called in the same callback, the node is removed regardless of replacement [source](./.skilld/pkg/README.md#thisremove)
- `this.skip()` is only available in `enter`, not `leave`, and skipping a node also prevents its `leave()` callback from firing -- this means ScopeTracker state can become inconsistent if you skip nodes that create scopes [source](./.skilld/issues/issue-106.md)
- Type-narrow `ScopeTrackerNode` using the `.type` discriminant or `instanceof` (v0.7.0+) to access parent declaration nodes -- `ScopeTrackerVariable` exposes `.variableNode`, `ScopeTrackerImport` exposes `.importNode`, `ScopeTrackerCatchParam` exposes `.catchNode`, `ScopeTrackerFunctionParam` exposes `.fnNode` [source](./.skilld/pkg/dist/index.d.ts:L268:311)
- `isCurrentScopeUnder(scope)` uses strict child checking -- it returns `false` when the current scope equals the given scope, only `true` for descendants [source](./.skilld/pkg/dist/index.d.ts:L226:236)
- Use `getUndeclaredIdentifiersInFunction(node)` to find free variables (closures) in a function or arrow function expression -- useful for tree-shaking analysis and determining if a function can be safely hoisted or moved [source](./.skilld/pkg/dist/index.d.ts:L246)
- `ScopeTrackerFunctionParam.start` and `.end` are deprecated -- access the function's position via `.fnNode.start` and `.fnNode.end` instead [source](./.skilld/pkg/dist/index.d.ts:L277:285)More Debugging skills
diagnosing-bugs
mattpocock/skills
Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.
explore-code
lllllllama/rigorpilot-skills
Rigor Improve implementation leaf skill for auditable candidate implementation in deep learning research repositories. Use when the researcher explicitly authorizes exploratory work on an isolated branch or worktree to transplant modules, adapt a backbone, add LoRA or adapter layers, replace a head, or stitch together meaningful low-risk migration ideas with rollback-aware records in `explore_outputs/`. Do not use for end-to-end exploration orchestration on top of `current_research`, trusted baseline reproduction, conservative debugging, environment setup, verified contribution claims, or default repository analysis.
safe-debug
lllllllama/rigorpilot-skills
Rigor Debug / Rigor Audit skill for deep learning research work. Use when the user pastes a traceback, terminal error, CUDA OOM, checkpoint load failure, shape mismatch, NaN loss symptom, or training failure and wants conservative diagnosis before any patching, with debug fixes clearly separated from research contributions. Do not use for broad refactoring, speculative adaptation, automatic exploratory patching, or general repository familiarization.

