Read Dokploy Logs

Unified log reader. Resolves the resource type and reads logs directly over MCP/REST (Dokploy v0.29.0+ — runtime logs are first-class over MCP/REST/CLI; no SSH/Beszel needed). For a multi-container compose stack, use /dokploy-dev:compose-logs instead — it reads every container.

Follow the read-logs skill (${CLAUDE_PLUGIN_ROOT}/skills/read-logs/SKILL.md) — load it first.

Arguments

Format: <resource-name-or-id> [--type <app|compose|db|deployment>] [--tail <N>] [--since <1h>] [--search <term>]

• resource-name-or-id — required. Name or id of the app / compose / database / deployment.
• --type — optional. Force a resource type. If omitted, auto-detect: application-one → compose-one → {type}-one (postgres/mysql/mariadb/mongo/redis/libsql) → deployment-allCentralized.
• --tail — recent lines, 1–10000 (default 200).
• --since — time window: all or <n>{s|m|h|d} (e.g. 30m, 2h). Default all.
• --search — server-side substring filter (e.g. error, ECONNREFUSED).

Parse from "$ARGUMENTS".

Process

1. Resolve the resource (see arg notes for the auto-detect chain).

2. Pick the right tool and read the log:

   Resource type	MCP call	Log kind
   Application	application-readLogs { applicationId, tail, since, search }	runtime stdout/stderr
   Compose	per-container — redirect to /dokploy-dev:compose-logs <name> (enumerate containers, then compose-readLogs { composeId, containerId, tail, since, search } for each)	runtime, per container
   Database (any of 6)	{type}-readLogs { {type}Id, tail, since, search }	runtime stdout/stderr
   Deployment	deployment-readLogs { deploymentId, tail } (find the id via deployment-all/deployment-allCentralized)	build log

   • The runtime-log .data is a newline-joined string, each line prefixed with an RFC3339 timestamp.
   • If a single compose resource is passed without /compose-logs, resolve composeId then read the first container but tell the user to run /dokploy-dev:compose-logs <name> to see all containers.

3. Format the output:

   • Show the last --tail lines.
   • Surface ERROR / FATAL / panic / exited / ECONNREFUSED / unauthorized lines first (see skill §6).
   • If a deployment is status: error, show the last 30 build-log lines up top.

4. Suggest the next step when an error pattern is found: /dokploy-dev:analyze (AI triage) or /dokploy-dev:debug <id> (full decision tree).

Example Usage

/dokploy-dev:logs web-frontend --since 1h
/dokploy-dev:logs web-frontend --tail 500 --search error
/dokploy-dev:logs pg-main --type db --tail 500
/dokploy-dev:logs <deploymentId> --type deployment
# multi-container stack → use the dedicated command:
/dokploy-dev:compose-logs my-stack