Read. Install.
Make progress.

Find, install, invoke.

Install the published GNCore Python package, verify the version you received, then choose skills for your assistant. The repository version is 4.0.0; the version available on PyPI may differ.

gncore search "clean data"
gncore list apps
gncore install sheet-clean --apps vscode-codex --scope workspace
gncore validate --apps vscode-codex

Then, in an assistant that supports native invocation, start a request with /sheet-clean followed by your dataset, context, constraints, and success criteria.

Workspace or user scope?

Workspace targets a specific project context where supported. User targets user-level reuse where supported. Use the --scope workspace or --scope user flag. Exact paths and skill-discovery behavior vary by assistant; use gncore list apps and inspect the relevant adapter.

Example: gncore install company-ceo --apps vscode-codex --scope user

Export one skill as a ZIP.

The catalog documents this command for hosted ChatGPT Skills:

gncore export --app chatgpt --skills company-ceo --output company-ceo.zip

It creates a per-skill ZIP for manual upload according to the documentation. Hosted ChatGPT support and file constraints are not equivalent to the twelve CLI installation targets; verify them against your package version and destination account.

Inspect before relying.

The published usage guide documents gncore doctor, gncore validate --apps vscode-codex, gncore backup --output gncore-backup.zip, gncore restore gncore-backup.zip, and gncore update --all. Review changes and version compatibility before restoring.

Skills guide work. They don't do it alone.

GNCore is reusable instruction content, not an autonomous software agent or a cloud-hosted AI execution service. Results depend on assistant functionality, tool permissions, provided context and human checks. Healthcare, legal and financial workflows are support for qualified review, not replacements for professional judgment.

Source verification status.

The public GNCore repository confirms the MIT license in its LICENSE file, and documents Python 3.10+ in pyproject.toml. Read the source usage guide for the current CLI.

This static website renders the published documentation and generated catalog; it does not host or execute GNCore Python code. Install GNCore from PyPI.