Skip to main content

Documentation Index

Fetch the complete documentation index at: https://filepacks.com/docs/llms.txt

Use this file to discover all available pages before exploring further.

​
What it does

pack walks an input directory, hashes each file, builds a format_version: 1 manifest, and writes a deterministic .fpk archive. Use it whenever a workflow finishes producing files and you want one portable artifact instead of a loose directory.

​
Usage

filepacks pack <input> --output <file>

​
When to use it

  • after an agent run writes output files
  • after an eval or CI job produces reports
  • before handing generated files to another reviewer or tool
  • before capturing a baseline for later comparison

​
Arguments

  • <input>: input directory to package
  • --output <file>: required output path for the .fpk artifact; it must end in .fpk
pack is the only public CLI command with a flag today, and --output is required.

​
Minimal example

npx filepacks pack ./run-output --output ./run-output.fpk

​
Expected output shape

Example: 
Pack
input=/absolute/path/to/input
output=/absolute/path/to/artifact.fpk
name=input
digest=sha256:...
files=2
bytes=12
The artifact name defaults to the input directory basename in the CLI.

​
Common mistakes

  • passing a file instead of a directory
  • forgetting --output
  • using an output path without the .fpk extension
  • trying to package symlinks or other non-regular files
  • assuming the CLI supports --name (it does not)
  • inspect to read the artifact summary you just created
  • verify before sharing or storing the artifact
  • compare to check the artifact against a baseline

​
Core API

@filepacks/core pack() accepts: 
await pack({
  input: '/tmp/input',
  output: '/tmp/example.fpk',
  name: 'example',
})
Unlike the CLI, the core API supports an optional name field.