Logs

DaoFlow provides three layers of observability for deployments.

Log Layers

Layer	Description	Format
Raw Logs	Exact stdout/stderr from containers	Plain text
Structured Timeline	Deploy steps with status and timing	JSON events
Agent Summaries	AI-readable failure analysis	Structured JSON

Viewing Logs

Via CLI

# Recent logs
daoflow logs --deployment dep_abc123 --lines 50

# Search failed readiness checks across recent control-plane deployments
daoflow logs control-plane --query readiness --stream stderr --lines 25

# Specific deployment
daoflow logs --deployment dep_abc123 --json

Via Dashboard

Navigate to Deployments → expand a deployment row to search persisted deployment logs. For live container logs, use Services → Logs.

Via API

GET /trpc/deploymentLogs?input={"json":{"deploymentId":"dep_abc123","query":"readiness","stream":"stderr"}}

Log Entry Schema

{
  "id": 1,
  "deploymentId": "dep_abc123",
  "level": "info",
  "message": "Container started successfully",
  "source": "runtime",
  "metadata": {},
  "createdAt": "2026-03-15T10:30:01Z"
}

Field	Values
level	debug, info, warn, error
source	build, runtime, system

Retention

• Raw logs are append-only and stored in PostgreSQL
• Structured events are normalized and queryable
• AI-generated summaries link back to exact log line IDs

External Log Drains

Operators can configure external log drains for webhook-style destinations, send a test payload, inspect delivery attempts, and retry failed deliveries.

daoflow log-drains list --json
daoflow log-drains create \
  --name ops-webhook \
  --type generic_http \
  --endpoint-url https://logs.example.com/ingest \
  --yes
daoflow log-drains deliveries --json

Delivery failures are stored with HTTP status, response body, and error text so operators can see what failed before retrying.

Required Scope

Reading deployment logs requires logs:read. Managing external log drains requires server:read for inventory and server:write for create, test, retry, and delete actions.