Observed since 20 Oct, 23:24 UTC
No superseded metadata states collected. Unchanged observations do not create new revisions.
Generic installer core for projects that distribute agent skills
agent-skill-installer installs agent skills for Codex and Claude Code from
local skill directories, GitHub repositories, or PyPI wheels. It supports
directory and global install targets, can assert directory targets are Git or
Sapling repositories with --repo, writes discoverability blocks into the agent
hook files, and records enough install state to safely upgrade or uninstall
skills it owns.
Additional agent targets or installer functionality are open for discussion, and pull requests are welcome.
python -m pip install agent-skill-installer
Run the interactive installer:
agent-skill-installer

Use --no-ui for scripts:
agent-skill-installer --no-ui install \
--skill-path ./my-skill \
--editable \
--agent codex \
--scope dir \
--repo
| Audience | Start here |
|---|---|
| Installing a skill | Installing Skills |
| Writing a skill directory or GitHub skill repo | Authoring Skills |
| Publishing a skill on PyPI or with platform wheels | Packaging |
| Embedding the installer API or exposing a wrapper CLI | API And Wrapper CLI |
| Maintaining releases | Maintainer Guide |
The examples/ directory contains runnable integrations:
examples/demo-installer/ is a complete Python package that carries a bundled
skill and exposes a project-specific wrapper command.examples/wheel-skill/ is a plain wheel-packaged skill for generic
--wheel-file installs.examples/api-install/ shows direct Python API usage without a wrapper
console script.Download counts and popularity will appear here as the collector receives them.
| Version | Published | Direct dependencies | Open version |
|---|---|---|---|
| 0.4.0 | 1 Sept 2026 | 3 |
This describes our collected data; it is not a security assessment.
Shared repository data · GitHub
Observed since 20 Oct, 23:24 UTC
No superseded metadata states collected. Unchanged observations do not create new revisions.
package:c77f4f7b01a6f4aaf65c7c28b030209d46b967e08dbfb6f46da5d05d9814eda4