rhdh-local
Operates a local Red Hat Developer Hub environment with the rhdh-local-setup customization system: enable or disable dynamic plugins, apply configuration, switch pristine and customized modes, start or stop containers, inspect health and logs, run plugin verification, and back up or restore customizations. Use for local RHDH, podman compose plugin testing, PR artifact verification, 504 or startup troubleshooting, and local Extensions Catalog checks.
How do I install this agent skill?
npx skills add https://github.com/redhat-developer/rhdh-skills --skill rhdh-localIs this agent skill safe to install?
- Gen Agent Trust Hubpass
The rhdh-local skill provides a secure local management system for Red Hat Developer Hub. It uses a standalone Python CLI to orchestrate container environments and synchronize configurations. The skill implements security best practices, including list-based subprocess execution to prevent command injection, robust path traversal protections for its backup and restore utilities, and explicit guardrails to prevent the exposure of sensitive credentials from environment files.
- Socketpass
No alerts
- Snykwarn
Risk: MEDIUM · 1 issue
What does this agent skill do?
RHDH Local
Own local execution and return evidence to the calling workflow. All commands use this skill's standalone CLI; another skill does not need to be installed.
Capability gate
- Run
uv run scripts/rhdh-local --help. - Locate
rhdh-local-setupthroughRHDH_LOCAL_SETUP_DIRor by walking from the current directory. Runuv run scripts/rhdh-local --json status. - If the setup is absent, state the required environment variable or checkout layout and stop only the runtime branch.
- Never read secrets from
.envinto conversation context. Ask only for missing variable names and let the container runtime consume their values.
Route by outcome
| Outcome | Load and follow |
|---|---|
| Read a plugin version or artifact at an overlay branch or tag | Run python3 scripts/fetch-plugin-metadata.py <plugin-name> --branch <branch-or-tag> --json; default ref is main and runtime setup is not required |
| Enable a catalog or PR artifact | workflows/enable-plugin.md |
| Disable a plugin | workflows/disable-plugin.md |
| Switch pristine/customized mode | workflows/switch-mode.md |
| Verify a plugin end to end | workflows/test-plugin.md and references/dynamic-plugin-testing.md |
| Inspect status or enabled packages | Run uv run scripts/rhdh-local --json status or uv run scripts/rhdh-local --json plugins list |
| Apply customizations | Run uv run scripts/rhdh-local --json apply |
| Start or stop | Run uv run scripts/rhdh-local --json up [flags] or uv run scripts/rhdh-local --json down |
| Check health | Run uv run scripts/rhdh-local --json health |
| Back up or restore | Run uv run scripts/rhdh-local --json backup or uv run scripts/rhdh-local restore <archive>; restore is a dry run until --force |
| Troubleshoot startup, networking, or 504s | references/troubleshooting.md |
| Configure environment variables | references/env-reference.md |
Invariants
- Edit only source files under
rhdh-customizations/. Runapplyafter every edit so copies underrhdh-local/stay synchronized. - Use this CLI's
upanddowncommands when Lightspeed or Orchestrator is enabled; they own the compose lifecycle and shared networks. - Obtain package references from
spec.dynamicArtifactor from the reference the caller supplied. Never construct OCI references from naming conventions. - Preserve the
includes:block in dynamic plugin overrides. Put backend packages before their frontend packages. - A plugin test succeeds only with recorded installation, boot, health, and UI results relevant to the request. Distinguish an expected credential error from a load failure.
What a caller supplies
A caller invokes this skill by name and states, in conversation:
- each package reference to enable, with its plugin config or
none - the environment variables the plugin needs, by name only
- the catalog test entities the plugin renders against
- which checks to run: installation, startup, health, UI
- whether to clean up afterwards
Use those references verbatim. Ask for a missing one rather than reconstructing it from a naming convention.
What this skill reports
Report to the caller in conversation. Do not write into another skill's directory.
- the subject tested, and the mode it ran in: customized or pristine
- one line per requested check — installation, startup, health, UI — each passed, failed, or skipped with the reason it was skipped
- the packages actually enabled, with their exact references
- the log excerpts that back each result
- cleanup state: completed, retained, or not requested
Completion
Complete when the report names every source customization file changed, every CLI command run, the health and UI evidence observed, the cleanup state, and one outcome for every check the caller requested.
How can the creator link this skill?
Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.
<a href="https://skillzs.dev/skills/redhat-developer/rhdh-skills/rhdh-local">View rhdh-local on skillZs</a>