nats-sinks 0.4.1

Production-ready outbound sink connectors for NATS JetStream.

nats-sinks

nats-sinks provides at-least-once delivery from JetStream to external destinations with commit-then-acknowledge processing and idempotent sink support.

The project repository is ProjectCuillin/nats-sinks. The current named contributor is Johan Louwers, reachable at [e-mail].

Overview

NATS is a lightweight messaging system used to move events between services. JetStream is the persistence layer in NATS: it stores messages in streams and delivers them to consumers. A sink is a consumer whose main job is to copy those messages into another durable system, such as Oracle Database, Oracle Autonomous Database on Oracle Cloud Infrastructure (OCI), or another approved storage backend.

nats-sinks is a Python package for building outbound NATS JetStream sink consumers. It provides a reusable runtime that owns JetStream delivery semantics and delegates destination writes to sink implementations. The current production sinks are Oracle Database, including OCI-hosted Oracle Autonomous Database deployments, Oracle MySQL, and local files.

The project is intentionally suitable for mission-oriented environments such as defence logistics, operational reporting, secure platform telemetry, and sensor-driven operational data flows where event loss, premature acknowledgement, and unclear audit trails are unacceptable. In defence and national-security settings, nats-sinks can be used as the durable event custody layer around command-and-control data fabrics, sensor-fusion pipelines, platform telemetry, weapon-system status events, sensor-to-shooter workflows, and kill-chain or kill-mesh style coordination messages. Its role is to preserve and persist operational events safely. It is not a targeting system, fire-control system, weapons-release mechanism, rules-of-engagement engine, or automation layer for lethal decision-making. The language throughout the documentation uses examples such as priority, classification, labels, DLQs, and encrypted payloads because those concepts map naturally to environments that must handle sensitive operational information with discipline.

The package is designed as a production-ready foundation rather than a demo script. It includes a typed public API, JSON configuration, a CLI, security-conscious defaults, tests, documentation, CI configuration, and packaging metadata suitable for publishing to PyPI.

The public documentation is prepared for Read the Docs at nats-sinks.readthedocs.io and a GitHub Pages mirror at projectcuillin.github.io/nats-sinks. Read the Docs is the preferred versioned documentation site for package users. GitHub Pages publishes the current main branch documentation after the repository Pages source is set to GitHub Actions.

Available Today

The current release is focused on a small production-ready surface that can be used immediately:

• JetStreamSinkRunner for pull-based JetStream consumption with bounded batches, commit-then-acknowledge processing, DLQ handling, graceful shutdown, logging hooks, basic metrics hooks, and safe redelivery behavior.
• NatsEnvelope, the immutable internal representation passed to sinks instead of raw NATS client messages.
• Core-normalized message metadata fields for priority, classification, and labels, with configurable NATS header extraction, defaults, and subject-specific rules shared by every sink. These fields are useful for separating routine traffic from urgent, restricted, coalition, exercise, or audit-relevant event streams without changing sink code.
• Optional data-centric security label profiles for structured releasability, handling caveats, owner, originator, policy identifier, and retention category metadata. Oracle stores the profile in SECURITY_LABELS_JSON, the file sink writes it as security_labels, and future sinks can use the same core-normalized profile.
• Optional message-level authenticity verification before any sink write, with subject-specific HMAC-SHA256 or Ed25519 signature rules, sanitized rejection reasons, and DLQ-before-ACK behavior for messages that fail verification. This complements NATS authentication and TLS by proving the message body and selected metadata were signed by an approved producer. See Message Authenticity.
• nats-sink, the CLI for validating JSON configuration, showing redacted effective config, testing sinks, running sink processes, and performing bounded read-only Oracle lineage queries by allow-listed mission metadata or message identity fields.
• nats-sink-metrics, a separate CLI for reading a local JSON metrics snapshot and rendering status as tables, JSON, JSONL, shell variables, metric names, or Prometheus text output.
• nats-sink-observe, a separate observability CLI for generating disabled sharing policies, validating what may be exported, and producing policy-filtered Prometheus textfile output for node_exporter or an optional native Prometheus HTTP scrape endpoint. It also provides a disabled-by-default OpenTelemetry OTLP metrics connector, Elastic Observability and Grafana Alloy profiles over the shared OTLP core, a disabled-by-default Splunk HEC connector for approved aggregate metrics, a disabled-by-default StatsD connector for best-effort datagram export, a disabled-by-default syslog bridge for bounded RFC 5424-style messages, and a disabled-by-default NATS server monitoring connector for explicitly approved /healthz, /jsz, and related endpoint fields.
• Optional core payload encryption for AES-256-GCM and AES-256-CCM before envelopes are delivered to Oracle, file, or future sinks.
• Optional tamper-evident custody metadata with deterministic payload, metadata, and record hashes computed before sink delivery.
• Optional pre-sink policy enforcement that runs after message normalization, metadata defaults, mission metadata validation, and payload encryption, but before any destination write. The policy gate can require priority, classification, labels, mission metadata, encrypted payloads, and payload size limits by subject. Rejected messages never reach a sink and follow the DLQ-before-ACK rule when DLQ is configured.
• Optional core size policy enforcement that can bound sink-bound payload bytes, normalized headers, labels, mission metadata, standard metadata, approximate record size, and accepted batch size before any sink write.
• nats_sinks.oracle.OracleSink, the production Oracle Database sink with connection pooling, Oracle Autonomous Database connection options, merge and insert_ignore idempotent modes, optional high-throughput staging-table merge mode, subject-to-table routing, metadata persistence, payload normalization, and explicit transaction commit before ACK.
• nats_sinks.mysql.MySqlSink, the production Oracle MySQL sink with connection pooling, TLS CA support, upsert and insert_ignore idempotent modes, subject-to-table routing, metadata persistence, payload normalization, and explicit transaction commit before ACK.
• nats_sinks.file.FileSink, the production local file sink with deterministic filenames, atomic temporary-file placement, optional fsync, duplicate handling, optional Python standard-library gzip compression, metadata persistence, and the same payload normalization contract used by Oracle.
• nats_sinks.spool.SpoolSink, the production-oriented encrypted edge spool sink for disconnected operation, bounded local custody, deterministic idempotency, priority-aware replay, and explicit replay into a final destination sink when connectivity returns.
• A safe sink connector framework with first-party Oracle Database, Oracle MySQL, file, and spool connectors, stable SinkConnector metadata, explicit SinkRegistry resolution, and disabled-by-default allow-listed entry-point discovery for reviewed external connectors.
• Basic metrics counters and timing observations for fetched, prepared, written, ACKed, NAKed, failed, DLQ, sink write, ACK error, active batch, and event freshness behavior. Freshness metrics cover event age at receive and store time, stale-event counts, missing or malformed creation timestamps, and positive source clock skew. The built-in runtime can write a local JSON snapshot when configured, and embedded applications can still supply their own metrics recorder or exporter. External observability sharing is controlled by a separate policy and is disabled by default, whether operators choose the recommended Prometheus textfile connector, the optional native HTTP endpoint, the OTLP connector, or NATS server monitoring.
• Optional JetStream advisory observation for selected $JS.EVENT.ADVISORY... subjects. Advisory support is disabled by default, produces aggregate metrics only, and never changes sink writes, retries, DLQ decisions, or JetStream ACK behavior.
• Explicit durable pull-consumer management with bind_only, create_if_missing, and reconcile modes so operators can choose whether the worker only binds to a pre-created consumer, creates a missing consumer, or reconciles compatible delivery settings before fetching messages.
• Richer durable consumer policy configuration for pull consumers, including plural filter subjects, server-side BackOff sequences, MaxDeliver, MaxAckPending, MaxWaiting, headers-only state, consumer replicas, memory-storage selection, and bounded low-sensitivity consumer metadata.
• NATS reconnect tuning for clustered or controlled-network deployments, including multiple seed URLs, reconnect wait, maximum reconnect attempts, ping behavior, pending buffer size, drain timeout, and connection event metrics.
• WebSocket NATS transport guardrails for approved ws:// local labs and wss:// deployments, including mixed transport rejection, credential-free URLs, local CA TLS handling, validated optional WebSocket headers, and a collision-safe local certification harness. See WebSocket Connection Evaluation.
• Exponential retry backoff with optional jitter for retryable failures, so temporary destination outages can slow down without weakening the commit-then-acknowledge invariant.
• Optional priority-aware processing lanes for already-fetched bounded batches, with weighted starvation controls, aggregate metrics, and explicit warnings that this is not exactly-once processing or strict total ordering. See Priority-Aware Processing Lanes.
• CycloneDX SBOM generation for release evidence, with JSON and XML SBOM files generated during local checks, CI builds, and release workflows.
• SHA-256 release checksum manifests and documented hash-verified installation guidance for high-trust deployment workflows.
• A deterministic synthetic mission scenario harness for generating fake NatsEnvelope test data and running local file-sink smoke checks without live NATS or Oracle services. See Synthetic Mission Testing.
• A use-case blueprint area that shows how generic features such as commit-then-ACK, mission metadata, classification, labels, payload encryption, Oracle storage, and file output can support mission-oriented patterns without making the product defence-only. See Use Cases.
• Kubernetes deployment examples that show JSON ConfigMaps, Secret references, mounted trust material, restrictive security contexts, resource limits, graceful termination, and optional Prometheus observability sidecars. See Kubernetes Deployment.
• A local Oracle Linux 9 slim based Docker image and JSON Compose stack for developer smoke testing with a temporary NATS JetStream service and the file sink. See Local Docker Stack.
• A local Oracle MySQL test database container for Oracle MySQL sink development and e2e testing, based on Oracle Linux 9 slim and Oracle MySQL 9.7.0 LTS, with random per-run credentials, loopback-only exposure, cleanup by default, and deterministic asset tests. See Oracle MySQL Test Container.
• A production container hardening baseline for the Oracle Linux slim image, including non-root UID/GID 10001, read-only-root-compatible runtime guidance, OCI image labels, writable-path documentation, SBOM and vulnerability-scanning expectations, and careful accreditation caveats. See Production Container Hardening.
• Mission-support operational examples that show complete patterns for restricted event storage, disconnected file handoff, DLQ triage and replay preparation, and destination outage recovery. See Mission-Support Operational Examples.
• A generic mission metadata profile for carrying validated JSON context to Oracle MISSION_METADATA_JSON, file-sink output records, and future sinks without adding fixed columns for every use case. See Mission Metadata.
• A data-centric security label profile for structured policy metadata such as releasability, handling caveats, owner, originator, policy ID, and retention category. See Data-Centric Security Label Profile.
• F2T2EA event phase tagging guidance built on mission metadata as metadata-only context, with explicit non-goals around targeting, fire-control, weapons-release, and autonomous decision behavior. See F2T2EA Event Phase Tagging.

Production sink modules shipped today:

• nats_sinks.oracle
• nats_sinks.mysql
• nats_sinks.file
• nats_sinks.spool

Status

The current release is 0.4.1.

Included today:

• Core JetStream pull-consumer runtime.
• Commit-then-acknowledge processing.
• Immutable NatsEnvelope abstraction.
• Explicit sink protocol and safe sink registry.
• Oracle sink with idempotent production modes.
• Oracle MySQL sink with idempotent production modes and container-backed e2e testing.
• File sink with atomic local JSON file writes and deterministic duplicate handling.
• Edge spool sink with encrypted local records and replay into a final sink.
• Optional AES-256-GCM and AES-256-CCM payload encryption in the core runner.
• Multi-key payload decryption helper for controlled key-rotation, replay, and verification workflows.
• Optional message authenticity verification for signed messages before payload encryption, policy checks, custody metadata, priority lanes, or sink writes.
• Optional tamper-evident custody metadata for deterministic verification evidence. Custody hashes are not encryption or digital signatures.
• JSON configuration and redacted effective-config output.
• CLI command named nats-sink.
• Metrics inspection command named nats-sink-metrics.
• Observability policy command named nats-sink-observe.
• Synthetic mission scenario harness for core and file-sink smoke testing.
• Generic mission metadata support with validated JSON context for Oracle, file sink, and future sinks.
• Optional pre-sink policy enforcement for fail-closed, destination-neutral checks before Oracle, file, or future sink writes.
• F2T2EA phase-tagging documentation as a use-case blueprint on top of mission metadata, not runtime workflow automation.
• Unit tests for ACK ordering, DLQ ordering, config loading, SQL generation, and Oracle mapping.
• Integration test placeholders isolated behind integration markers.
• MkDocs documentation, examples, GitHub Actions workflows, governance files, and security policy.

The package does not claim exactly-once delivery. It provides at-least-once delivery with clear commit ordering and idempotent sink support. That means a message may be delivered more than once, especially after failures, and sinks must be configured so duplicate processing is safe.

Architecture

flowchart LR
    Producer[Publisher] --> Stream[JetStream stream]
    Stream --> Consumer[Durable pull consumer]
    Consumer --> Runner[nats-sinks core runner]
    Runner --> Envelope[NatsEnvelope batch]
    Envelope --> Auth{authenticity required?}
    Auth -->|passes or disabled| Crypto{payload encryption enabled?}
    Auth -. permanent rejection .-> DLQ
    Crypto -->|yes| Encrypted[Encrypted payload envelope]
    Crypto -->|no| Plain[Original payload bytes]
    Encrypted --> Policy{pre-sink policy enabled?}
    Plain --> Policy
    Policy -->|passes| Custody{custody metadata enabled?}
    Policy -. permanent rejection .-> DLQ
    Custody --> Sink[sink.write_batch]
    Sink --> Commit[Durable destination commit]
    Commit --> Ack[JetStream ACK]

    Runner -. permanent failure .-> DLQ[DLQ publish]
    DLQ -. publish succeeds .-> Ack

The core rule is:

Core owns delivery semantics. Sinks own destination writes.

Sinks never receive raw NATS messages. Sinks never ACK messages. The core runtime converts raw messages into NatsEnvelope instances, calls the sink, and ACKs only after durable success.

Commit-Then-Acknowledge

The project invariant is:

A JetStream message must only be acknowledged after all required durable side effects have completed successfully. ACK is the final confirmation of successful processing, never a prerequisite for processing.

Short slogan:

Commit first. ACK last. Design for redelivery.

The normal processing sequence is:

sequenceDiagram
    participant JS as JetStream
    participant R as JetStreamSinkRunner
    participant S as Destination Sink
    participant D as Durable Destination

    JS->>R: Deliver message batch
    R->>R: Normalize into NatsEnvelope
    R->>S: write_batch(envelopes)
    S->>D: Write rows or records
    D-->>S: Commit succeeds
    S-->>R: Return success
    R->>JS: ACK messages

If the sink fails before durable commit, the core does not ACK. If commit succeeds but the process exits before ACK, JetStream may redeliver the message. That is expected and must be handled by idempotency.

Installation

pip install nats-sinks
pip install "nats-sinks[oracle]"
pip install "nats-sinks[crypto]"
pip install "nats-sinks[dev]"
pip install "nats-sinks[docs]"
pip install "nats-sinks[all]"

Python >=3.11 is required.

Quick Start

Start NATS with JetStream:

nats-server -js -m 8222
nats stream add ORDERS --subjects "orders.*"
nats pub orders.created '{"order_id":"O-1001","amount":42.50}'

Prepare the destination:

For a local no-database quick start, use the file sink. It writes one JSON file per message under .local/file-sink/events, which is ignored by git. See File Sink for the full configuration and durability model.

Run the sink:

nats-sink validate examples/file-basic/config.json
nats-sink test-sink examples/file-basic/config.json
nats-sink run examples/file-basic/config.json

For a mission-system prototype, the file sink is often the fastest way to prove the delivery contract before connecting a database. It preserves payloads and metadata in ordinary JSON files so operators and maintainers can inspect the flow, confirm classification and label handling, and validate redelivery behavior without needing database access.

JSON Configuration

Runtime configuration is JSON-only. The package uses the standard-library JSON parser for application configuration. The generic nats, delivery, dead_letter, logging, and metrics sections are shared by all sinks. The sink object selects the destination and contains destination-specific fields documented on each sink page.

{
  "nats": {
    "url": "nats://localhost:4222",
    "urls": [],
    "stream": "ORDERS",
    "consumer": "file-orders-sink",
    "subject": "orders.*",
    "durable": true,
    "no_echo": false,
    "allow_reconnect": true,
    "reconnect_time_wait_seconds": 2,
    "max_reconnect_attempts": 60
  },
  "delivery": {
    "batch_size": 100,
    "batch_timeout_ms": 1000,
    "max_in_flight_batches": 1,
    "ack_policy": "after_sink_commit",
    "max_retries": 5,
    "retry_backoff_ms": 1000,
    "retry_backoff_max_ms": 60000,
    "retry_backoff_mode": "exponential",
    "retry_backoff_multiplier": 2.0,
    "retry_jitter": "full",
    "prefer_safe_duplication": true
  },
  "dead_letter": {
    "enabled": true,
    "subject": "orders.dlq",
    "include_payload": true,
    "include_headers": true,
    "include_error": true,
    "ack_term_after_publish": false
  },
  "logging": {
    "level": "INFO",
    "payload_logging": false
  },
  "metrics": {
    "enabled": false,
    "namespace": "nats_sinks",
    "snapshot_file": null,
    "event_freshness_enabled": true,
    "event_stale_after_seconds": 300,
    "event_future_skew_tolerance_seconds": 5
  },
  "message_metadata": {
    "priority": {
      "header": "Nats-Sinks-Priority",
      "default": "normal"
    },
    "classification": {
      "header": "Nats-Sinks-Classification",
      "default": null
    }
  },
  "encryption": {
    "enabled": false,
    "algorithm": "aes-256-gcm",
    "key_id": "orders-runtime-key",
    "key_b64_env": "NATS_SINKS_PAYLOAD_KEY_B64"
  },
  "sink": {
    "type": "file",
    "directory": ".local/file-sink/events",
    "filename_strategy": "stream_sequence",
    "duplicate_policy": "skip_existing",
    "payload_mode": "json_or_envelope",
    "fsync": true
  }
}

Secret values should come from the environment or a secret manager. Use environment-backed fields such as password_env and token_env rather than storing credentials in config files.

To write a local dependency-free metrics snapshot, enable metrics and set a snapshot path:

{
  "metrics": {
    "enabled": true,
    "namespace": "nats_sinks",
    "snapshot_file": ".local/nats-sinks/metrics.json",
    "event_freshness_enabled": true,
    "event_stale_after_seconds": 300,
    "event_future_skew_tolerance_seconds": 5
  }
}

Inspect the snapshot from another terminal:

nats-sink-metrics show .local/nats-sinks/metrics.json --format table
nats-sink-metrics show .local/nats-sinks/metrics.json --format shell --kind counter
nats-sink-metrics show .local/nats-sinks/metrics.json --metric "oracle_*"
nats-sink-metrics show .local/nats-sinks/metrics.json --metric "mysql_*"
nats-sink-metrics show .local/nats-sinks/metrics.json --metric "event_*" --metric "events_*"
nats-sink-metrics get .local/nats-sinks/metrics.json messages_failed_total --default 0

The metrics CLI is documented in Metrics. Policy-controlled Prometheus and OpenTelemetry export are part of the observability documentation, including the Elastic Observability and Grafana Alloy profiles, Splunk HEC connector, StatsD connector, and syslog bridge. Start with Observability, then use Prometheus Integration, OpenTelemetry OTLP Integration, Elastic Observability Profile, or Grafana Alloy Profile, or Splunk HEC Integration, or StatsD Integration, or Syslog Bridge for connector details. The NATS server monitoring connector and delivery-boundary decision for endpoints such as /jsz and /healthz are documented in NATS Server Monitoring. Optional JetStream advisory counters are configured through Configuration and documented in Metrics. Durable consumer startup behavior is documented under consumer_management.

Payload Bodies

NATS message bodies are bytes. The generic framework accepts bytes and does not require JSON at the core boundary. Sinks that store data in JSON-capable destinations can use the shared payload normalization contract for JSON, encrypted text, plain text, and opaque bytes.

The default payload_mode is json_or_envelope:

• standards-compliant JSON is stored unchanged,
• non-JSON UTF-8 text is wrapped in a JSON envelope,
• non-text bytes are wrapped as base64 in the same JSON envelope.

Python-only JSON constants such as NaN, Infinity, and -Infinity are not treated as valid JSON. In json_or_envelope mode they are preserved as text in the payload envelope; in json_only mode they are permanent serialization failures that can follow the configured DLQ path.

For encrypted text streams where the ciphertext may or may not decrypt to JSON later, use payload_mode: "text_envelope" to wrap every body as text and avoid unnecessary JSON parse attempts.

{
  "sink": {
    "type": "file",
    "directory": ".local/file-sink/events",
    "payload_mode": "text_envelope"
  }
}

See Sink Framework and File Sink for the JSON envelope shape and operational guidance. Oracle-specific payload storage is documented in Oracle Sink.

Message Authenticity

The core runner can verify producer-level message signatures before the envelope reaches any sink. This is different from NATS authentication and TLS: those controls protect broker access and network transport, while message authenticity proves that the message body and selected metadata were signed by an approved producer key.

Supported algorithms are HMAC-SHA256 and Ed25519. Rules can apply to every subject or to selected NATS wildcard subject families:

{
  "message_authenticity": {
    "enabled": true,
    "unmatched_subject_action": "reject",
    "rules": [
      {
        "subject": "mission.>",
        "algorithm": "hmac-sha256",
        "key_id": "producer-key-2026-05",
        "key_b64_env": "NATS_SINKS_AUTHENTICITY_KEY_B64",
        "signed_fields": ["subject", "message_id"]
      }
    ]
  }
}

Verification failures are permanent pre-sink failures. The message never reaches Oracle, file, spool, or future sinks. If DLQ is configured, the original JetStream message is ACKed only after the DLQ publish succeeds.

See Message Authenticity for the canonical signed document, producer examples, Ed25519 guidance, metrics, and DLQ behavior.

Payload Encryption

The core runner can encrypt the message body before sending an envelope to any sink. This protects the actual payload stored by Oracle, file, and future sinks, while leaving operational metadata such as subject, headers, stream sequence, message IDs, and timestamps readable for routing and idempotency.

Supported algorithms are AES-256-GCM and AES-256-CCM through the optional nats-sinks[crypto] extra. Encryption can apply to every subject consumed by the runner or to selected subjects through ordered NATS wildcard rules:

{
  "encryption": {
    "enabled": true,
    "algorithm": "aes-256-gcm",
    "key_id": "orders-prod-2026-05",
    "key_b64_env": "NATS_SINKS_PAYLOAD_KEY_B64"
  }
}

For subject-specific encryption, leave the global policy disabled and add rules. The first matching rule wins; subjects with no matching rule remain unchanged in this example:

{
  "encryption": {
    "enabled": false,
    "rules": [
      {
        "subject": "secure.>",
        "enabled": true,
        "algorithm": "aes-256-gcm",
        "key_id": "secure-prod-2026-05",
        "key_b64_env": "NATS_SINKS_SECURE_PAYLOAD_KEY_B64"
      }
    ]
  }
}

Use stable metadata-based idempotency such as stream sequence or message ID when encryption is enabled. Ciphertext is intentionally non-deterministic because each encryption uses a fresh nonce.

See Payload Encryption for the full configuration reference, encrypted JSON envelope shape, testing script, and decryption helper.

Metadata Capture

nats-sinks captures a generic metadata JSON document for every message. This is available to all current and future sinks through NatsEnvelope.

The metadata document preserves all message headers, known NATS-reserved headers when present, unknown future Nats- headers, JetStream stream and sequence metadata, optional reply subject, and timing fields. Optional headers such as Nats-Msg-Id or Nats-Expected-Stream may be absent; that is normal and does not cause a crash. Destination sinks can store this document directly or map selected fields into destination-specific columns.

The core also normalizes three application-level metadata fields on every message: priority, classification, and labels. They can be supplied by NATS headers such as Nats-Sinks-Priority, Nats-Sinks-Classification, and Nats-Sinks-Labels, configured with deployment defaults, configured with ordered subject-specific defaults, or left unset. Headers always win when present; subject defaults are used only when the corresponding header is absent. Missing priority and classification values are stored as JSON null or SQL NULL, not as the literal string "null". Labels are normalized as a list and are stored in scalar sink fields as semicolon-separated text.

Classification and priority values are operator-defined strings. The documentation uses NATO-style examples such as NATO UNCLASSIFIED, NATO RESTRICTED, NATO CONFIDENTIAL, NATO SECRET, and COSMIC TOP SECRET; use the exact vocabulary required by your environment.

{
  "message_metadata": {
    "priority": {
      "header": "Nats-Sinks-Priority",
      "default": "routine"
    },
    "classification": {
      "header": "Nats-Sinks-Classification",
      "default": "NATO UNCLASSIFIED"
    },
    "labels": {
      "header": "Nats-Sinks-Labels",
      "default": "logistics;default"
    },
    "rules": [
      {
        "subject": "mission.reports.>",
        "priority": "immediate",
        "classification": "NATO SECRET",
        "labels": "mission-report;coalition;watch-floor"
      }
    ]
  }
}

With the file sink, these values appear as top-level JSON fields such as "priority": "immediate", "classification": "NATO SECRET", "labels": "mission-report;coalition;watch-floor", and "labels_list": ["mission-report", "coalition", "watch-floor"]. With Oracle, the same values are stored in PRIORITY, CLASSIFICATION, and LABELS columns and repeated inside METADATA_JSON.message_metadata.

For richer data-centric policy context, enable security_labels. The profile is optional and metadata-only. It can help downstream systems reason about releasability, handling caveats, owner, originator, policy identifiers, and retention categories, but it does not replace authorization in Oracle, IAM, or the destination platform.

{
  "security_labels": {
    "enabled": true,
    "allowed_classifications": [
      "NATO UNCLASSIFIED",
      "NATO RESTRICTED",
      "NATO CONFIDENTIAL",
      "NATO SECRET"
    ],
    "default": {
      "profile": "nats-sinks.security-label.v1",
      "classification": "NATO RESTRICTED",
      "releasability": ["NATO"],
      "handling_caveats": ["MISSION"],
      "owner": "example-owner",
      "originator": "example-originator",
      "policy_id": "example-policy",
      "retention_category": "mission-log-30d"
    }
  }
}

With the file sink, the profile appears as top-level security_labels and inside metadata.security_labels. With Oracle, the profile is stored in SECURITY_LABELS_JSON and repeated inside METADATA_JSON.security_labels.

NATS Connections

nats-sinks supports common NATS client authentication options through the nats JSON section:

• token authentication with token_env or token,
• username/password authentication with user and password_env or password,
• server-side bcrypted username/password credentials using the same client-side user and password_env settings,
• NATS credentials-file and decentralized JWT user workflows through creds_file,
• NKEY challenge authentication through nkey_seed_file,
• TLS server verification with tls_ca_file, including private or self-signed NATS server CAs,
• TLS client certificate/key transport settings for mutual TLS deployments,
• optional no_echo connection behavior for reviewed same-connection publish/subscribe policies.

Do not embed credentials in nats.url; use environment-backed fields instead. See NATS Connections And Authentication for configuration examples and secure deployment notes.

Authentication is only half of the production story. The NATS runtime account should also have least-privilege subject permissions: fetch from the configured pull consumer, receive inbox replies, ACK received messages after durable sink success, and publish to the configured DLQ subject only when DLQ is enabled. See NATS Least-Privilege Permissions for templates and validation checklists.

For stream setup review, nats-sink stream-plan can generate an offline JetStream planning report for retention, discard policy, storage, replicas, duplicate-window settings, and runtime versus administration permissions. It does not connect to NATS or modify stream state. See JetStream Stream Management Planning.

CLI

nats-sink --help
nats-sink validate examples/file-basic/config.json
nats-sink test-sink examples/file-basic/config.json
nats-sink validate examples/oracle-jetstream/config.json
nats-sink show-effective-config examples/oracle-jetstream/config.json
nats-sink stream-plan examples/oracle-jetstream/config.json
nats-sink query-lineage examples/oracle-jetstream/config.json --field mission_id --value MISSION-ALPHA --dry-run
nats-sink test-sink examples/oracle-jetstream/config.json
nats-sink run examples/oracle-jetstream/config.json
nats-sink-metrics show .local/nats-sinks/metrics.json --format table
nats-sink-metrics show .local/nats-sinks/metrics.json --metric "oracle_*"
nats-sink-metrics show .local/nats-sinks/metrics.json --metric "event_*" --metric "events_*"
nats-sink-metrics get .local/nats-sinks/metrics.json messages_failed_total --default 0
nats-sink-observe init-prometheus-policy examples/file-basic/config.json .local/observability.prometheus.json
nats-sink-observe validate-policy .local/observability.prometheus.json

The CLI:

• returns non-zero on validation or runtime errors,
• prints the active sink type,
• prints the commit-then-acknowledge ACK policy,
• renders effective configuration as redacted JSON,
• never prints resolved passwords.

Read-only Oracle lineage query helpers are documented in Lineage Query Helpers.

The metrics CLI reads only a local JSON snapshot written by the runner when metrics.enabled and metrics.snapshot_file are configured. It supports table, JSON, JSONL, shell, names, and Prometheus text output so developers can pipe results into service checks and scripts. Oracle Database duplicate/conflict counters are visible through the same command with --metric "oracle_*", and Oracle MySQL duplicate/upsert counters are visible with --metric "mysql_*". Event freshness and staleness metrics are visible with --metric "event_*" and --metric "events_*" when metrics are enabled. See Metrics for examples.

The observability CLI manages external sharing policy. It can generate a disabled Prometheus policy from runtime config, list known metric names and subject hints, validate the policy, write policy-filtered Prometheus textfile output, run a disabled-by-default native Prometheus HTTP endpoint, and export approved metrics to an OpenTelemetry Collector through OTLP/HTTP JSON, including Elastic Observability and Grafana Alloy profiles that reuse the shared OTLP core, approved aggregate metric export to Splunk HEC, and best-effort StatsD datagram and syslog message export. Metrics sharing remains off until the global policy and the selected connector are explicitly enabled. See Observability, Prometheus Integration, OpenTelemetry OTLP Integration, and the Elastic Observability Profile or Grafana Alloy Profile or Splunk HEC Integration or StatsD Integration or Syslog Bridge for connector guidance.

Python API

You can use nats-sinks directly from another Python project without shelling out to the CLI. The recommended integration point is the public framework API:

from nats_sinks import JetStreamSinkRunner
from nats_sinks.file import FileSink

sink = FileSink(
    directory="/var/lib/nats-sinks/events",
    filename_strategy="stream_sequence",
    duplicate_policy="skip_existing",
)

runner = JetStreamSinkRunner(
    nats_url="nats://localhost:4222",
    stream="ORDERS",
    consumer="orders-file-sink",
    subject="orders.*",
    sink=sink,
)

await runner.run()

You can also mount the Typer CLI into another Typer application:

import typer
from nats_sinks.cli.main import app as nats_sink_cli

app = typer.Typer()
app.add_typer(nats_sink_cli, name="nats-sink")

See Python Usage for embedded application patterns and the tradeoff between using the public runtime API and importing CLI internals.

Documented imports and CLI entry points are protected by public API compatibility tests. See Public API Compatibility for the supported import contract and release guard.

Production Sinks

Destination-specific details are split into dedicated pages:

• Oracle Sink covers Oracle connection types, Autonomous Database, table DDL, least-privilege users, idempotent write modes, subject-to-table routing, payload storage, metadata columns, and Oracle-specific performance guidance.
• Oracle MySQL Sink covers Oracle MySQL connection settings, TLS CA files, recommended table DDL, least-privilege users, idempotent upsert and insert_ignore modes, subject-to-table routing, payload storage, metadata columns, and the local container-backed e2e test.
• File Sink covers local file output, atomic write behavior, deterministic file names, duplicate policies, gzip compression, filesystem safety, and file-specific performance guidance.
• Edge Spool Sink covers encrypted local custody for disconnected operation, bounded spool directories, deterministic duplicate handling, priority-aware replay, and forwarding into a final destination sink.

The generic sink framework is documented separately in Sink Framework and the reusable release gate is documented in Sink Certification. That boundary is deliberate: Oracle Database, Oracle MySQL, file, and spool sinks use the same core delivery semantics, the same envelope contract, and the same commit-then-acknowledge rule. Future sinks must provide comparable certification evidence before they are described as production-ready.

Generic data-handling features such as payload encryption, tamper-evident custody metadata, mission metadata, and priority lanes are documented separately from sink-specific pages so new sinks can adopt them without changing the public delivery contract.

Failure Behavior

sequenceDiagram
    participant JS as JetStream
    participant R as Runner
    participant S as Sink
    participant DLQ as DLQ subject

    JS->>R: Deliver invalid message
    R->>S: write_batch(envelope)
    S-->>R: PermanentSinkError
    R->>DLQ: Publish diagnostic JSON
    DLQ-->>R: Publish acknowledged
    R->>JS: ACK original message

Important failure cases:

• destination write or commit fails: no ACK, message redelivers according to the JetStream consumer policy,
• destination commit succeeds and the process crashes before ACK: message may redeliver, so sink idempotency must handle the duplicate,
• payload is permanently invalid for the selected sink: message is published to DLQ when configured, then the original is ACKed only after DLQ publish succeeds,
• message authenticity verification rejects a missing, malformed, mismatched, or invalid signature: the message does not reach a sink, and the original is ACKed only after DLQ publish succeeds when DLQ is configured,
• DLQ publish fails: original message is not ACKed.

Security Notes

• Do not store secrets in repository files.
• Do not log payloads by default.
• Do not log passwords, tokens, private keys, NATS credentials, Oracle credentials, or full connection strings.
• SQL identifiers are allow-list validated.
• SQL values use bind variables.
• Unit tests must not make network calls.
• Integration tests are isolated behind markers.
• Use TLS and authenticated NATS connections in production.
• Use least-privilege NATS permissions for the sink runtime account; avoid broad publish, broad subscribe, stream administration, and source-subject publish rights for ordinary workers.
• Use core payload encryption when destination storage should retain encrypted message bodies while keeping routing metadata available.
• Use message authenticity verification when broker authentication and TLS are not enough to prove producer-level event provenance.
• Use tamper-evident custody metadata when operators need deterministic hashes for later verification, and remember that hashes are not encryption.
• Use least-privilege destination credentials with access only to the required destination resources.

Development

python -m pip install -e ".[dev,oracle,mysql,crypto,docs]"
ruff format --check .
ruff check .
mypy src
pytest
python -m build
python scripts/update-dependency-manifests.py --check
scripts/sbom.sh
twine check dist/*.whl dist/*.tar.gz

Run only unit tests:

pytest -m "not integration"

Build documentation:

scripts/check-docs.sh

The docs helper builds the Read the Docs and GitHub Pages variants in isolated temporary directories. That prevents overlapping MkDocs runs from cleaning the same site/ directory.

Manual live NATS connection testing is documented in NATS Connections And Authentication and Testing. The tracked helper script is scripts/nats-live-probe.py; real CA files and credentials should stay under ignored .local/ paths.

The latest sanitized validation summary is maintained in Latest Test Report. That report is overwritten in place for each new validation run and must not contain server addresses, usernames, passwords, tokens, certificate contents, wallet material, connection strings, or sensitive payloads.

To run nats-sink as a systemd service on Oracle Linux or Debian, see Service Deployment. The repository includes example service files under examples/systemd/ and a unified OS-detecting installer at scripts/install-systemd.sh. The documented one-command installer downloads the script from ProjectCuillin/nats-sinks and runs it with sudo. When the script is not running inside a checkout, it fetches the required example configuration files and systemd unit files from the selected GitHub ref. Production operators should pin a release tag and inspect the script first when policy requires reviewed installation steps.

Kubernetes deployment guidance is documented in Kubernetes Deployment. The tracked examples under examples/kubernetes/ are public-safe starting points that operators must customize before use.

Release and PyPI publishing instructions are documented in Publishing Releases. That guide covers version updates, tag pushes, GitHub release workflows, TestPyPI, PyPI trusted publishing, and manual fallback commands.

Development now follows a branch-first release workflow. Maintainers should create release-*, feature-*, bugfix-*, or hotfix-* branches, push small changes to those branches, and merge to main only through reviewed pull requests. See Branch-First Development And Release Workflow for the quiet-branch policy, manual release validation, branch protection, pull request, and release-tag rules.

Backlog and feature-request workflow is documented in Backlog Management. GitHub Issues are the live backlog; CHANGELOG.md records work that has shipped or is staged for the next release. Maintainers can define local backlog items as JSON files under backlog/items/ and sync them to GitHub Issues with scripts/sync-backlog-issues.py or the Backlog Sync GitHub Actions workflow. The backlog tooling also validates target-release labels and rejects common public-leak patterns such as network locators, IP literals, credential assignments, token-like values, and certificate blocks before content is posted to GitHub Issues. Issue lifecycle tooling supports the standard maintainer flow: assign the issue, post a sanitized implementation plan, apply the release label, post sanitized test-plan and close-out evidence, tick acceptance criteria, and leave the issue open until release automation closes it after the associated GitHub Release exists.

Dependency inventory guidance is documented in Dependency Management. pyproject.toml is the authoritative package metadata, and generated requirements*.txt manifests are kept in sync for GitHub Dependency Graph, Dependabot, and dependency review workflows.

Hash-verified installation guidance for controlled deployment environments is documented in Hash-Verified Installs.

Repository Layout

src/nats_sinks/core      Core runtime, config, envelope, runner, DLQ
src/nats_sinks/sinks     Sink protocols and registry
src/nats_sinks/oracle    Oracle sink implementation
src/nats_sinks/mysql     Oracle MySQL sink implementation
src/nats_sinks/file      Local file sink implementation
src/nats_sinks/cli       CLI entry point
tests/unit               Deterministic unit tests
tests/integration        External-service and local end-to-end tests
docs                     MkDocs documentation
examples                 Local development examples

Roadmap

Future work is intentionally listed near the end of the README so new readers first see what the package can do today. Planned items are not production features until they are implemented, tested, documented, and released.

Phase 1:

• Core runtime.
• Oracle sink.
• Oracle MySQL sink.
• File sink.
• NATS reconnect tuning and connection event metrics.
• WebSocket connection guardrails, optional headers, and local certification harness.
• Policy-controlled Prometheus textfile export and optional native Prometheus HTTP scrape endpoint as separate observability services.
• Policy-controlled OpenTelemetry OTLP metrics export to an OpenTelemetry Collector as a separate observability command or service.
• Disabled-by-default Elastic Observability and Grafana Alloy profiles over the shared OTLP observability core.
• Disabled-by-default Splunk HEC connector for approved aggregate metrics in security operations and incident-response environments.
• Disabled-by-default StatsD connector for approved best-effort UDP or Unix datagram metric export.
• Disabled-by-default syslog bridge for approved bounded RFC 5424-style metric messages over UDP or Unix datagram sockets.
• Disabled-by-default NATS server monitoring connector for approved endpoint fields, implemented outside the delivery worker.
• Kubernetes deployment examples with worker/observability separation, security contexts, resource limits, graceful shutdown settings, and Secret references.
• Least-privilege NATS permissions templates for runtime workers, DLQ publish rights, optional consumer management, and advisory readers.
• Offline JetStream stream-management planning helper for retention, discard, storage, replicas, duplicate-window, and permission review.
• Advanced JetStream topology guidance for mirrors, sources, subject transforms, republish behavior, stream compression, placement, metadata, and idempotency review.
• CycloneDX SBOM generation for release evidence.
• CLI.
• Documentation.
• Tests.
• PyPI-ready package.

Phase 2:

• More idempotency strategies.
• First-party Oracle-family sink designs for OCI Object Storage, Oracle Berkeley DB, Oracle NoSQL Database, and OCI Streaming.
• High-priority Palantir Foundry and Palantir Gotham connector evaluations with local contract mocks before any live certification claim.
• Additional Oracle MySQL HeatWave tuning and certification guidance.
• HTTP sink.
• S3 sink design with deterministic object keys.
• Native OCI Object Storage sink design with deterministic object keys, workload identity support, checksums, multipart upload, and least-privilege bucket guidance.
• Kafka, search, warehouse, document database, key-value, and wide-column backend evaluation through the sink framework.
• Public Docker image release publication automation, including signed image publication and container-specific provenance attachments.
• Expanded live certification runbooks for NATS TLS certificate, NKEY, and decentralized JWT deployments across representative server policies.
• Deeper sequence-based and timestamp-based JetStream replay-start controls.
• Optional confirmed ACK and InProgress support.
• Payload-presence metadata and sink certification for headers-only delivery.

Phase 3:

• External connector marketplace guidance and certification evidence beyond the current allow-listed entry-point framework.
• Sink certification tests.
• Helm chart.
• Additional advanced observability connectors and bounded subject-aware metric policies.
• Push and ordered consumer evaluation where compatible with project semantics.
• Stream management helpers beyond the current topology guidance.
• Future sink certification tests.

Not planned unless scope changes:

• AckNone, early ACK, and AckAll behavior that weakens commit-then-ack.
• General-purpose Core NATS pub/sub, queue group, request/reply, or services framework support.
• JetStream Key/Value and Object Store APIs unless a future sink needs them.

See NATS Feature Gap Analysis for the detailed comparison.

SBOM generation and release evidence are documented in SBOM And Release Evidence. Release asset checksums and hash-verified installation workflows are documented in Hash-Verified Installs.

License

Apache-2.0. See LICENSE.