fix(build): replace bun build with bun run build

Author: gustavovalverde

.github/workflows/performance-regression.yml

@@ -41,7 +41,7 @@ jobs:
         run: bun install --frozen-lockfile
 
       - name: Build project
-        run: bun build
+        run: bun run build
 
       - name: Restore main baseline
         run: |
@@ -140,7 +140,7 @@ jobs:
         run: bun install --frozen-lockfile
 
       - name: Build project
-        run: bun build
+        run: bun run build
 
       - name: Capture baseline
         run: node bench/capture-baseline.js

.github/workflows/release.yml

@@ -47,7 +47,7 @@ jobs:
           bun lint
           bun typecheck
           bun test
-          bun build
+          bun run build
 
       - name: Create Release Pull Request or Publish
         id: changesets

.github/workflows/test.yml

@@ -47,4 +47,4 @@ jobs:
         run: bun test
 
       - name: Build
-        run: bun build
+        run: bun run build

CONTRIBUTING.md

@@ -22,7 +22,7 @@ Thanks for your interest in improving h2m-parser! Keeping cognitive load low is
 - Husky runs `lint-staged` to format/lint staged files. Make sure your commits are clean.
 - Use [Changesets](https://github.com/changesets/changesets) for release notes. After feature work, run `bun changeset` to document changes and version bumps.
 - When the release PR opens, review the changelog and, after merging, approve the pending run in the `release` environment to publish.
-- CI runs `bun lint`, `bun typecheck`, `bun test`, and `bun build`. Keep the `verify` script green locally before pushing.
+- CI runs `bun lint`, `bun typecheck`, `bun test`, and `bun run build`. Keep the `verify` script green locally before pushing.
 
 ## Reporting issues
 
@@ -54,7 +54,7 @@ bun regression
 bun qa
 
 # Development watch mode
-bun build:watch    # Terminal 1
+bun run build:watch    # Terminal 1
 bun test:watch     # Terminal 2
 ```
 
@@ -93,7 +93,7 @@ bun test:watch     # Terminal 2
 
 ```bash
 # Start development
-bun build:watch     # In terminal 1
+bun run build:watch     # In terminal 1
 bun test:watch     # In terminal 2
 
 # After making changes

bench/README.md

@@ -5,7 +5,7 @@ This folder contains the benchmarking harness described in `performance-implemen
 ## Quick start
 
 ```bash
-bun build
+bun run build
 bun bench quick
 ```
 

bench/utils/h2m-loader.js

@@ -27,7 +27,7 @@ async function ensureDistBuilt() {
         if (code === 0) {
           resolve();
         } else {
-          reject(new Error(`bun build exited with code ${code}`));
+          reject(new Error(`bun run build exited with code ${code}`));
         }
       });
     }).finally(() => {

docs/PERFORMANCE-REGRESSION.md

@@ -179,7 +179,7 @@ bun bench:compare:full
 
 ```bash
 # Watch mode - rerun on file changes
-bun build:watch & bun bench:compare:quick
+bun run build:watch & bun bench:compare:quick
 ```
 
 > ℹ️ **Configuration parity:** `bun bench:baseline`, `bun bench:regression`, and `bun bench:readme --fresh` all invoke the same comparison run (100 iterations, 3 warmups, 5MB per-file limit, dataset `tests/fixtures`). This keeps the baseline, regression gate, and published metrics directly comparable.

docs/RELEASE.md

@@ -33,7 +33,7 @@ Need to kick a release outside of an automatic merge? Use **Actions → Release
 
 - *Publish fails with auth errors*: confirm the npm trusted publisher entry still points at `.github/workflows/release.yml` and the `release` environment exists. Re-run the job after fixing.
 - *No release PR appears*: make sure there is at least one unpublished changeset file on `main`; otherwise the workflow exits early. You can run it manually via `workflow_dispatch`.
-- *Need to bypass automation*: as a last resort you can run `bun verify && bun build && npm publish --access public`, but follow up by re-running `bun changeset version` so history stays in sync.
+- *Need to bypass automation*: as a last resort you can run `bun verify && bun run build && npm publish --access public`, but follow up by re-running `bun changeset version` so history stays in sync.
 
 ## Retired workflows
 

docs/SCRIPTS.md

@@ -4,11 +4,11 @@ This document mirrors the scripts defined in `package.json` so newcomers can dis
 
 ## Build & Packaging
 
-- `bun build` – Build the distributable bundles in `dist/` via tsup.
-- `bun build:watch` – Rebuild bundles on file changes (handy while working on the CLI).
+- `bun run build` – Build the distributable bundles in `dist/`.
+- `bun run build:watch` – Rebuild bundles on file changes (handy while working on the CLI).
 - `bun build:binary` – Compile standalone executable for current platform.
 - `bun build:binary:all` – Build binaries for all platforms (Linux, macOS, Windows)
-- `bun prepack` – Convenience alias used by npm when publishing; runs `bun build`.
+- `bun prepack` – Convenience alias used by npm when publishing; runs `bun run build`.
 
 ## Quality Gates
 
@@ -27,12 +27,12 @@ This document mirrors the scripts defined in `package.json` so newcomers can dis
 ## Publishing
 
 - `bun changeset` – Create or edit Changesets.
-- `bun release` – Build and publish using Changesets automation (calls `bun build` internally).
+- `bun release` – Build and publish using Changesets automation (calls `bun run build` internally).
 - `bun prepare` – Husky hook installer (runs automatically after install).
 
 ## Benchmarking & Performance Tooling
 
-The benchmarking utilities live under `bench/` and expect a fresh build (`bun build`) before running because they import from the compiled bundles. The `bun bench` command opens an interactive menu that wraps the most common flows; every entry has a matching script for CI usage.
+The benchmarking utilities live under `bench/` and expect a fresh build (`bun run build`) before running because they import from the compiled bundles. The `bun bench` command opens an interactive menu that wraps the most common flows; every entry has a matching script for CI usage.
 
 ### Everyday commands
 
@@ -77,7 +77,7 @@ The benchmarking utilities live under `bench/` and expect a fresh build (`bun bu
 ```bash
 bun install        # bootstrap
 bun verify         # lint + typecheck + tests
-bun build          # produce dist/ bundles
+bun run build      # produce dist/ bundles
 bun bench          # open interactive benchmark menu
 bun bench quick    # run quick benchmark via CLI shortcut
 bun bench:compare  # compare h2m-parser vs. other converters (Markdown output)

examples/README.md

@@ -67,7 +67,7 @@ Make sure you've built the project first:
 bun install
 
 # Build the library
-bun build
+bun run build
 
 # Run an example
 node examples/basic.js
@@ -130,4 +130,4 @@ const converter = new H2MParser({
 
 - Check the [main README](../README.md) for detailed API documentation
 - Report issues on [GitHub](https://github.com/gustavovalverde/h2m-parser/issues)
-- See [CONTRIBUTING.md](../CONTRIBUTING.md) for development guidelines
+- See [CONTRIBUTING.md](../CONTRIBUTING.md) for development guidelines