NoETL Tools Reference (Canonical v10)

NoETL tools are executed as tasks inside a canonical step pipeline (step.tool).

Canonical v10 reminders:

• A step is: admission policy (step.spec.policy.admit) + ordered task pipeline (step.tool) + router (step.next arcs).
• Task outcome handling is done via task.spec.policy.rules (no legacy eval/expr).
• Large outputs are reference-first (ResultRef/ManifestRef patterns).

See documentation/docs/reference/dsl/step_spec.md for the full DSL model.

---

Available tool kinds

Kind	Description	Use case
http	HTTP requests	APIs, webhooks, polling
postgres	PostgreSQL commands/queries	OLTP + result storage
python	Inline Python execution	transforms, custom logic
duckdb	DuckDB queries	analytics, local joins
snowflake	Snowflake queries	warehouse operations
container	Job/container execution	heavy deps, isolation
gcs	Google Cloud Storage	export/import artifacts
ducklake	Lakehouse queries	unified analytics
nats	JetStream/KV/Object Store	state, messaging, artifacts

---

Canonical usage pattern

Basic task in a pipeline

- step: fetch_ping
  tool:
    - call:
        kind: http
        method: GET
        url: "{{ workload.api_url }}/ping"
        spec:
          policy:
            rules:
              - when: "{{ outcome.status == 'error' }}"
                then: { do: fail }
              - else:
                  then: { do: break }

With routing

next:
  spec: { mode: exclusive }
  arcs:
    - step: end
      when: "{{ event.name == 'step.done' }}"

With authentication

Canonical approach:

• declare required credentials under root keychain
• reference credentials by name in tasks (for example auth: pg_k8s)

See documentation/docs/reference/auth_and_keychain_reference.md.

---

Outcome envelope (canonical)

Each task produces an outcome:

• outcome.status: "ok" or "error"
• outcome.result: success output (small inline payload or reference)
• outcome.error: structured error (kind required; retryable recommended)
• outcome.meta: timing/attempt metadata

Policies (task.spec.policy.rules) evaluate over outcome.

---

Template namespaces

Common namespaces available in templates:

• workload.* (immutable merged input)
• keychain.* (resolved credentials; read-only)
• ctx.* (execution-scoped mutable context)
• iter.* (iteration-scoped mutable context; loops only)
• args.* (token payload / arc inscription)
• pipeline locals: _prev, _task, _attempt, outcome (policy evaluation)
• routing: event (boundary event; next.arcs[].when)

---

See also

• documentation/docs/reference/dsl/spec.md
• documentation/docs/reference/dsl/runtime_events.md
• documentation/docs/reference/dsl/runtime_results.md
• documentation/docs/reference/retry_mechanism_v2.md
• documentation/docs/reference/pagination_v2.md