trunkio-launcher-skilld

name	trunkio-launcher-skilld
description	ALWAYS use when writing code importing "@trunkio/launcher". Consult for debugging, best practices, or modifying @trunkio/launcher, trunkio/launcher, trunkio launcher.
metadata	{"version":"1.3.4","generated_by":"Anthropic · Haiku 4.5","generated_at":"2026-07-01T00:00:00.000Z"}

@trunkio/launcher@1.3.4

Tags: latest: 1.3.4

References: package.json • README • Docs

Search

Use skilld search "query" -p @trunkio/launcher instead of grepping .skilld/ directories. Run skilld search --guide -p @trunkio/launcher for full syntax, filters, and operators.

API Changes

@trunkio/launcher is a CLI bootstrap tool, not a library with versioned public APIs. It serves as a minimal npm wrapper that downloads and runs the Trunk CLI binary for the pinned version in your trunk.yaml configuration.

No documented public APIs or version-specific changes. The launcher itself has no exported functions, classes, or composables — it's invoked exclusively via the trunk CLI command. Version changes in the launcher reflect underlying Trunk CLI binary updates, not API surface changes.

For Trunk CLI feature changes, consult the changelog in the Trunk platform documentation, which tracks new features, fixes, and deprecations in the Trunk product itself (merge queue, flaky tests, etc.).

Best Practices

• Commit .trunk/trunk.yaml to version control with pinned CLI version — ensures all developers and CI use the same Trunk version, eliminating "works on my machine" issues with linters and formatters source

• Store SHA256 checksums in .trunk/trunk.yaml under cli.sha256.<platform> — enables security verification of downloaded Trunk binaries before execution source

• Use TRUNK_CLI_VERSION environment variable to override pinned version only in exceptional circumstances — allows temporary version testing without committing changes, but defaults to .trunk/trunk.yaml for reproducibility source

• Configure TRUNK_CACHE environment variable to use a shared cache directory on CI systems — reduces repeated binary downloads across parallel CI jobs and improves pipeline speed source

• Verify binary integrity after download by checking SHA256 sum — protects against corrupted downloads or man-in-the-middle attacks during CI runs source

• Run trunk upgrade in a dedicated commit to bump version in .trunk/trunk.yaml — keeps version changes visible in git history and allows the team to review Trunk version updates separately source

• Place .trunk/trunk.yaml at repo root and search directory trees from project subdirectories — enables Trunk to work correctly in monorepos and nested project structures source

• Rely on automatic platform detection for binary downloads — the launcher handles downloading the correct prebuilt binary for macOS, Linux, and Windows without manual configuration source

• Forward all CLI arguments directly to the Trunk executable — ensures the launcher is transparent and supports all Trunk CLI features without wrapper overhead source

• Use the launcher as a drop-in replacement for a native Trunk installation — no additional configuration needed beyond installing @trunkio/launcher via npm and setting up .trunk/trunk.yaml source

• Leverage automatic caching in ~/.cache/trunk (or XDG_CACHE_HOME/trunk) — cached binaries are reused across multiple runs and projects, reducing disk I/O and network traffic source

• Validate .trunk/trunk.yaml syntax with a YAML parser when committing — prevents silent failures if YAML is malformed, since parsing errors will only occur during Trunk execution source

• Never commit .trunk/tools or cache directories — these contain downloaded binaries and should be regenerated per environment, not checked into version control source