From azzist
Generate azzist config files for a project: a committed azzist.yaml skeleton and a gitignored azzist.local.yaml for secrets (Cloudflare token, SSH credentials). Use when the user starts an azzist deployment, says "azzist init", "set up azzist config", or before any azzist scaffold/deploy step when no azzist.yaml exists yet.
How this skill is triggered — by the user, by Claude, or both
Slash command
/azzist:azzist-initThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Create the two-file config split so secrets never land in a public repo.
Create the two-file config split so secrets never land in a public repo.
This skill is idempotent and only concerns config files — it creates or updates
azzist.yaml + azzist.local.yaml, nothing else.
azzist.yaml (project name, stack, port, db/cache, repo, branch). Don't re-ask for
things analyze already figured out — just confirm with the user once.azzist.yaml already exists AND azzist.local.yaml already exists with the expected
keys filled in, STOP and tell the user: "already initialized — azzist.yaml at
and azzist.local.yaml (gitignored) are in place. Re-run azzist-analyze to refresh
detection, or edit the files directly." Don't overwrite them silently..gitignore excludes the secrets file. If .gitignore does not
already contain azzist.local.yaml, append these lines:
azzist.local.yaml
**/azzist.local.yaml
Verify with git check-ignore azzist.local.yaml (should print the path) before
continuing. If the project is not a git repo, still create .gitignore so it is safe
the moment it becomes one.azzist.yaml:
${CLAUDE_PLUGIN_ROOT}/templates/config/azzist.yamlazzist.local.yaml:
${CLAUDE_PLUGIN_ROOT}/templates/config/azzist.local.yaml.exampleapi_token + zone_id, and per-server host, ssh_user, auth mode.ssh_config alias > key > password. Only use password
if the user insists, and warn it needs sshpass locally.azzist.yaml is committed (no secrets).azzist.local.yaml is gitignored (holds secrets) — never commit it.Report which files were created and exactly which fields the user still needs to fill in
azzist.local.yaml before deploy can run.
Provides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
npx claudepluginhub ghulammuzz/azzist-skills --plugin azzist