Stats
Actions
Tags
From coolify-audit
Read-only audit of a Coolify instance — full resource inventory, env-var hygiene (plaintext secrets), exposed ports, and servers that need validation. Writes NOTHING. Use when the user says "/coolify-audit", "audit my coolify", "review coolify config", "any exposed ports in coolify", "check coolify secrets", "audita mi coolify", "revisa la configuración de coolify", "¿hay puertos expuestos en coolify?", "revisa los secretos de coolify". Requires scripts/coolify.py + configured creds (token needs `read:sensitive` for the env-value checks to be meaningful).
How this skill is triggered — by the user, by Claude, or both
Slash command
/coolify-audit:coolify-auditThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Never writes. Reply in the user's language (EN/ES). `CO = scripts/coolify.py`.
Never writes. Reply in the user's language (EN/ES). CO = scripts/coolify.py.
python $CO resources --json → everything on the instance.python $CO apps list, db list, svc list, server list.python $CO env list <uuid> --target <app|db|svc> --json.KEY=, TOKEN=,
PASSWORD=, private keys). NOTE: values are masked unless the token has read:sensitive
— if they come back empty, say so rather than reporting "no secrets".apps get <uuid> read ports_exposes / ports_mappings; list anything publicly mapped.fqdn is intended.python $CO server list → for any unvalidated server, note its uuid and tell the
user/operator to run python $CO server validate <uuid> themselves (or hand off to
the coolify skill). Do NOT run server validate here — it triggers a server-side
validation pipeline and is therefore a side effect, not a read.--apply. If a fix is needed, hand off to the coolify skill.Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub jachana/claude-marketplace --plugin coolify-audit