audit-core/audit_core/sqlite_backend.py

403 lines
15 KiB
Python
Raw Permalink Normal View History

Route ingestion through the backend contract; fix error semantics AUDIT-WP-0004 T01, T02, T07. T01 - ingestion wrote to SQLite directly and never called the AuditBackend contract, so a 202 meant a row existed rather than that a backend with a declared retention policy had accepted the event. Adds IdempotentAuditBackend to the contract: duplicate detection lives inside the backend so custody and idempotency state share a transaction and cannot diverge. SQLiteAuditBackend implements it with WAL, synchronous=FULL and a busy timeout. Ingestion now refuses any backend declaring durable=False, so the development file backend cannot silently become the production sink. The atomicity claim was tested rather than asserted, and the first attempt failed: with a single shared connection, 16 racing submissions of one event told two callers they were first. Storage was correct but the response was not. Fixed with per-thread connections and BEGIN IMMEDIATE around the insert/read pair, and locked in by a test. T02 - storage errors previously escaped the handler with start_response never called, and the auth check sat outside the try block so a non-ASCII Authorization header crashed the request. Adds a catch-all, maps conflict to 409, backend unavailability to 503 and unexpected faults to 500, and documents the full response contract with the retry semantics each status implies, since senders key their behaviour off it. T07 - ingestion tests 2 -> 23, suite 15 -> 36. accepted_at is now UTC rather than local time, and naive timestamps are rejected instead of silently assumed. Remaining in WP-0004: T03 tenant/source binding, T04 redaction policy, T05 operator read surface, T06 production serving layer. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-10 14:30:20 +02:00
"""Durable SQLite audit backend.
Implements the idempotent backend contract for single-node deployments and for
development. Production custody moves to PostgreSQL under AUDIT-WP-0005; this
backend stays the development and test implementation and defines the
behaviour the Postgres backend must match.
"""
from __future__ import annotations
import json
import sqlite3
import threading
from datetime import datetime, timezone
from audit_core.interface import (
AcceptResult,
AuditEvent,
BackendUnavailableError,
EventConflictError,
EventValidationError,
RetentionPolicy,
validate_event,
)
from audit_core.integrity import (
GENESIS,
ChainRow,
attestation_from_report,
chain_link,
verify_rows,
)
Route ingestion through the backend contract; fix error semantics AUDIT-WP-0004 T01, T02, T07. T01 - ingestion wrote to SQLite directly and never called the AuditBackend contract, so a 202 meant a row existed rather than that a backend with a declared retention policy had accepted the event. Adds IdempotentAuditBackend to the contract: duplicate detection lives inside the backend so custody and idempotency state share a transaction and cannot diverge. SQLiteAuditBackend implements it with WAL, synchronous=FULL and a busy timeout. Ingestion now refuses any backend declaring durable=False, so the development file backend cannot silently become the production sink. The atomicity claim was tested rather than asserted, and the first attempt failed: with a single shared connection, 16 racing submissions of one event told two callers they were first. Storage was correct but the response was not. Fixed with per-thread connections and BEGIN IMMEDIATE around the insert/read pair, and locked in by a test. T02 - storage errors previously escaped the handler with start_response never called, and the auth check sat outside the try block so a non-ASCII Authorization header crashed the request. Adds a catch-all, maps conflict to 409, backend unavailability to 503 and unexpected faults to 500, and documents the full response contract with the retry semantics each status implies, since senders key their behaviour off it. T07 - ingestion tests 2 -> 23, suite 15 -> 36. accepted_at is now UTC rather than local time, and naive timestamps are rejected instead of silently assumed. Remaining in WP-0004: T03 tenant/source binding, T04 redaction policy, T05 operator read surface, T06 production serving layer. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-10 14:30:20 +02:00
_SCHEMA = """
CREATE TABLE IF NOT EXISTS events (
event_id TEXT PRIMARY KEY,
payload_hash TEXT NOT NULL,
accepted_at TEXT NOT NULL,
correlation_id TEXT,
tenant TEXT NOT NULL,
record TEXT NOT NULL,
chain_hash TEXT,
chain_prev TEXT
Route ingestion through the backend contract; fix error semantics AUDIT-WP-0004 T01, T02, T07. T01 - ingestion wrote to SQLite directly and never called the AuditBackend contract, so a 202 meant a row existed rather than that a backend with a declared retention policy had accepted the event. Adds IdempotentAuditBackend to the contract: duplicate detection lives inside the backend so custody and idempotency state share a transaction and cannot diverge. SQLiteAuditBackend implements it with WAL, synchronous=FULL and a busy timeout. Ingestion now refuses any backend declaring durable=False, so the development file backend cannot silently become the production sink. The atomicity claim was tested rather than asserted, and the first attempt failed: with a single shared connection, 16 racing submissions of one event told two callers they were first. Storage was correct but the response was not. Fixed with per-thread connections and BEGIN IMMEDIATE around the insert/read pair, and locked in by a test. T02 - storage errors previously escaped the handler with start_response never called, and the auth check sat outside the try block so a non-ASCII Authorization header crashed the request. Adds a catch-all, maps conflict to 409, backend unavailability to 503 and unexpected faults to 500, and documents the full response contract with the retry semantics each status implies, since senders key their behaviour off it. T07 - ingestion tests 2 -> 23, suite 15 -> 36. accepted_at is now UTC rather than local time, and naive timestamps are rejected instead of silently assumed. Remaining in WP-0004: T03 tenant/source binding, T04 redaction policy, T05 operator read surface, T06 production serving layer. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-10 14:30:20 +02:00
);
CREATE INDEX IF NOT EXISTS events_correlation_idx ON events (correlation_id);
CREATE INDEX IF NOT EXISTS events_tenant_idx ON events (tenant);
Bind sender identities, add operator read surface, real serving layer AUDIT-WP-0004 T03, T05, T06. T03 - the receiver accepted whatever tenant and source a caller sent as long as it held the one shared token, despite WP-0003 recording tenant isolation as delivered. audit_core.senders binds each credential to the sources and tenants it may assert, driven by AUDIT_CORE_SENDERS rather than literals. Identities hold a list of tokens so rotation publishes the replacement alongside the incumbent and needs no delivery gap. Read is a separate privilege from write, so a sender credential cannot read the audit trail back. T05 - lookup by event id, lookup by correlation id, and a dead-letter view. Rejections are recorded rather than silently dropped. An event rejected for carrying secret-shaped material has its payload withheld: storing it would write that material into the audit store, which is what the rejection exists to prevent. Reason and payload hash are kept so it stays traceable. Replay is deliberately not built here. Idempotent replay is a property of the durable store and building it against SQLite would produce a second implementation to throw away; it lands with the Postgres backend in AUDIT-WP-0005-T01. T06 - serving moves to waitress with configurable threads and channel timeout, installed in the image via the serve extra. Without it the entrypoint falls back to a threaded wsgiref server with a socket timeout and graceful shutdown on SIGTERM, and logs a warning so a deployment cannot quietly land on the fallback. Metric counters deferred to WP-0005-T03 to be designed against the real scrape path. Tests 36 -> 46, covering cross-tenant and cross-source refusal, token rotation, read/write privilege separation, correlation lookup, and payload withholding on secret rejection. Remaining in WP-0004: T04 redaction policy, which needs a decision on whether a secret-shaped field is a rejection or a redaction. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-10 14:50:02 +02:00
CREATE TABLE IF NOT EXISTS dead_letters (
id INTEGER PRIMARY KEY AUTOINCREMENT,
event_id TEXT,
received_at TEXT NOT NULL,
sender TEXT,
reason TEXT NOT NULL,
payload_hash TEXT NOT NULL,
payload TEXT,
payload_withheld INTEGER NOT NULL DEFAULT 0
);
CREATE INDEX IF NOT EXISTS dead_letters_event_idx ON dead_letters (event_id);
2026-08-10 16:02:22 +02:00
-- Counted per field path, not merely per event: the point is to stop senders
-- emitting secret-shaped fields, and that needs the offending path named.
CREATE TABLE IF NOT EXISTS secret_findings (
sender TEXT NOT NULL,
source TEXT NOT NULL,
action TEXT NOT NULL,
field_path TEXT NOT NULL,
outcome TEXT NOT NULL,
persisted INTEGER NOT NULL DEFAULT 0,
occurrences INTEGER NOT NULL DEFAULT 0,
first_seen TEXT NOT NULL,
last_seen TEXT NOT NULL,
PRIMARY KEY (sender, source, action, field_path, outcome)
);
Route ingestion through the backend contract; fix error semantics AUDIT-WP-0004 T01, T02, T07. T01 - ingestion wrote to SQLite directly and never called the AuditBackend contract, so a 202 meant a row existed rather than that a backend with a declared retention policy had accepted the event. Adds IdempotentAuditBackend to the contract: duplicate detection lives inside the backend so custody and idempotency state share a transaction and cannot diverge. SQLiteAuditBackend implements it with WAL, synchronous=FULL and a busy timeout. Ingestion now refuses any backend declaring durable=False, so the development file backend cannot silently become the production sink. The atomicity claim was tested rather than asserted, and the first attempt failed: with a single shared connection, 16 racing submissions of one event told two callers they were first. Storage was correct but the response was not. Fixed with per-thread connections and BEGIN IMMEDIATE around the insert/read pair, and locked in by a test. T02 - storage errors previously escaped the handler with start_response never called, and the auth check sat outside the try block so a non-ASCII Authorization header crashed the request. Adds a catch-all, maps conflict to 409, backend unavailability to 503 and unexpected faults to 500, and documents the full response contract with the retry semantics each status implies, since senders key their behaviour off it. T07 - ingestion tests 2 -> 23, suite 15 -> 36. accepted_at is now UTC rather than local time, and naive timestamps are rejected instead of silently assumed. Remaining in WP-0004: T03 tenant/source binding, T04 redaction policy, T05 operator read surface, T06 production serving layer. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-10 14:30:20 +02:00
"""
Bind sender identities, add operator read surface, real serving layer AUDIT-WP-0004 T03, T05, T06. T03 - the receiver accepted whatever tenant and source a caller sent as long as it held the one shared token, despite WP-0003 recording tenant isolation as delivered. audit_core.senders binds each credential to the sources and tenants it may assert, driven by AUDIT_CORE_SENDERS rather than literals. Identities hold a list of tokens so rotation publishes the replacement alongside the incumbent and needs no delivery gap. Read is a separate privilege from write, so a sender credential cannot read the audit trail back. T05 - lookup by event id, lookup by correlation id, and a dead-letter view. Rejections are recorded rather than silently dropped. An event rejected for carrying secret-shaped material has its payload withheld: storing it would write that material into the audit store, which is what the rejection exists to prevent. Reason and payload hash are kept so it stays traceable. Replay is deliberately not built here. Idempotent replay is a property of the durable store and building it against SQLite would produce a second implementation to throw away; it lands with the Postgres backend in AUDIT-WP-0005-T01. T06 - serving moves to waitress with configurable threads and channel timeout, installed in the image via the serve extra. Without it the entrypoint falls back to a threaded wsgiref server with a socket timeout and graceful shutdown on SIGTERM, and logs a warning so a deployment cannot quietly land on the fallback. Metric counters deferred to WP-0005-T03 to be designed against the real scrape path. Tests 36 -> 46, covering cross-tenant and cross-source refusal, token rotation, read/write privilege separation, correlation lookup, and payload withholding on secret rejection. Remaining in WP-0004: T04 redaction policy, which needs a decision on whether a secret-shaped field is a rejection or a redaction. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-10 14:50:02 +02:00
# Rejection reasons whose payload must never be persisted. Storing the body of
# an event rejected *for containing secret-shaped material* would write that
# material into the audit store — the precise outcome the rejection prevents.
WITHHOLD_PAYLOAD_REASONS = frozenset({"secret_shaped_field"})
Route ingestion through the backend contract; fix error semantics AUDIT-WP-0004 T01, T02, T07. T01 - ingestion wrote to SQLite directly and never called the AuditBackend contract, so a 202 meant a row existed rather than that a backend with a declared retention policy had accepted the event. Adds IdempotentAuditBackend to the contract: duplicate detection lives inside the backend so custody and idempotency state share a transaction and cannot diverge. SQLiteAuditBackend implements it with WAL, synchronous=FULL and a busy timeout. Ingestion now refuses any backend declaring durable=False, so the development file backend cannot silently become the production sink. The atomicity claim was tested rather than asserted, and the first attempt failed: with a single shared connection, 16 racing submissions of one event told two callers they were first. Storage was correct but the response was not. Fixed with per-thread connections and BEGIN IMMEDIATE around the insert/read pair, and locked in by a test. T02 - storage errors previously escaped the handler with start_response never called, and the auth check sat outside the try block so a non-ASCII Authorization header crashed the request. Adds a catch-all, maps conflict to 409, backend unavailability to 503 and unexpected faults to 500, and documents the full response contract with the retry semantics each status implies, since senders key their behaviour off it. T07 - ingestion tests 2 -> 23, suite 15 -> 36. accepted_at is now UTC rather than local time, and naive timestamps are rejected instead of silently assumed. Remaining in WP-0004: T03 tenant/source binding, T04 redaction policy, T05 operator read surface, T06 production serving layer. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-10 14:30:20 +02:00
class SQLiteAuditBackend:
"""Store audit events in SQLite with idempotent accept semantics.
Configured for durability rather than speed: WAL journalling, ``synchronous
= FULL`` so an acknowledged write has reached disk, and a busy timeout so
concurrent writers wait instead of raising immediately.
"""
def __init__(self, path: str, retention_days: int | None = None, busy_timeout_ms: int = 5000) -> None:
self.path = path
self.retention_days = retention_days
self.busy_timeout_ms = int(busy_timeout_ms)
# One connection per thread. A shared connection lets concurrent
# statements interleave, which was observed to let two callers both
# believe they were the first to accept the same event.
self._local = threading.local()
with self._connect_raw() as setup:
setup.executescript(_SCHEMA)
self._ensure_chain_columns(setup)
Route ingestion through the backend contract; fix error semantics AUDIT-WP-0004 T01, T02, T07. T01 - ingestion wrote to SQLite directly and never called the AuditBackend contract, so a 202 meant a row existed rather than that a backend with a declared retention policy had accepted the event. Adds IdempotentAuditBackend to the contract: duplicate detection lives inside the backend so custody and idempotency state share a transaction and cannot diverge. SQLiteAuditBackend implements it with WAL, synchronous=FULL and a busy timeout. Ingestion now refuses any backend declaring durable=False, so the development file backend cannot silently become the production sink. The atomicity claim was tested rather than asserted, and the first attempt failed: with a single shared connection, 16 racing submissions of one event told two callers they were first. Storage was correct but the response was not. Fixed with per-thread connections and BEGIN IMMEDIATE around the insert/read pair, and locked in by a test. T02 - storage errors previously escaped the handler with start_response never called, and the auth check sat outside the try block so a non-ASCII Authorization header crashed the request. Adds a catch-all, maps conflict to 409, backend unavailability to 503 and unexpected faults to 500, and documents the full response contract with the retry semantics each status implies, since senders key their behaviour off it. T07 - ingestion tests 2 -> 23, suite 15 -> 36. accepted_at is now UTC rather than local time, and naive timestamps are rejected instead of silently assumed. Remaining in WP-0004: T03 tenant/source binding, T04 redaction policy, T05 operator read surface, T06 production serving layer. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-10 14:30:20 +02:00
def _connect_raw(self) -> sqlite3.Connection:
try:
db = sqlite3.connect(self.path, isolation_level=None)
db.execute("PRAGMA journal_mode = WAL")
db.execute("PRAGMA synchronous = FULL")
db.execute(f"PRAGMA busy_timeout = {self.busy_timeout_ms}")
return db
except sqlite3.Error as exc:
raise BackendUnavailableError(f"cannot open audit store: {exc}") from exc
@property
def db(self) -> sqlite3.Connection:
conn = getattr(self._local, "conn", None)
if conn is None:
conn = self._local.conn = self._connect_raw()
return conn
@property
def retention_policy(self) -> RetentionPolicy:
return RetentionPolicy(
custody_class="development",
retention_days=self.retention_days,
immutable=False,
tamper_evidence=False,
durable=True,
)
def emit(self, event: AuditEvent) -> str:
"""Persist an event, generating no idempotency guarantee."""
return self.accept(event, payload_hash=_record_hash(event)).reference
def accept(self, event: AuditEvent, payload_hash: str) -> AcceptResult:
try:
validate_event(event)
except EventValidationError:
raise
reference = f"audit:{event.event_id}"
details = event.details if isinstance(event.details, dict) else {}
db = self.db
try:
# BEGIN IMMEDIATE takes the write lock up front, so the insert and
# the follow-up read are one atomic pair. Without it, two callers
# racing on the same event id can both be told they were first.
db.execute("BEGIN IMMEDIATE")
except sqlite3.Error as exc:
raise BackendUnavailableError(str(exc)) from exc
try:
head = db.execute(
"SELECT chain_hash FROM events "
"ORDER BY accepted_at DESC, event_id DESC LIMIT 1"
).fetchone()
previous = head[0] if head and head[0] else GENESIS
link = chain_link(previous, payload_hash, event.event_id)
Route ingestion through the backend contract; fix error semantics AUDIT-WP-0004 T01, T02, T07. T01 - ingestion wrote to SQLite directly and never called the AuditBackend contract, so a 202 meant a row existed rather than that a backend with a declared retention policy had accepted the event. Adds IdempotentAuditBackend to the contract: duplicate detection lives inside the backend so custody and idempotency state share a transaction and cannot diverge. SQLiteAuditBackend implements it with WAL, synchronous=FULL and a busy timeout. Ingestion now refuses any backend declaring durable=False, so the development file backend cannot silently become the production sink. The atomicity claim was tested rather than asserted, and the first attempt failed: with a single shared connection, 16 racing submissions of one event told two callers they were first. Storage was correct but the response was not. Fixed with per-thread connections and BEGIN IMMEDIATE around the insert/read pair, and locked in by a test. T02 - storage errors previously escaped the handler with start_response never called, and the auth check sat outside the try block so a non-ASCII Authorization header crashed the request. Adds a catch-all, maps conflict to 409, backend unavailability to 503 and unexpected faults to 500, and documents the full response contract with the retry semantics each status implies, since senders key their behaviour off it. T07 - ingestion tests 2 -> 23, suite 15 -> 36. accepted_at is now UTC rather than local time, and naive timestamps are rejected instead of silently assumed. Remaining in WP-0004: T03 tenant/source binding, T04 redaction policy, T05 operator read surface, T06 production serving layer. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-10 14:30:20 +02:00
inserted = db.execute(
"""
INSERT INTO events
(event_id, payload_hash, accepted_at, correlation_id, tenant, record,
chain_hash, chain_prev)
VALUES (?, ?, ?, ?, ?, ?, ?, ?)
Route ingestion through the backend contract; fix error semantics AUDIT-WP-0004 T01, T02, T07. T01 - ingestion wrote to SQLite directly and never called the AuditBackend contract, so a 202 meant a row existed rather than that a backend with a declared retention policy had accepted the event. Adds IdempotentAuditBackend to the contract: duplicate detection lives inside the backend so custody and idempotency state share a transaction and cannot diverge. SQLiteAuditBackend implements it with WAL, synchronous=FULL and a busy timeout. Ingestion now refuses any backend declaring durable=False, so the development file backend cannot silently become the production sink. The atomicity claim was tested rather than asserted, and the first attempt failed: with a single shared connection, 16 racing submissions of one event told two callers they were first. Storage was correct but the response was not. Fixed with per-thread connections and BEGIN IMMEDIATE around the insert/read pair, and locked in by a test. T02 - storage errors previously escaped the handler with start_response never called, and the auth check sat outside the try block so a non-ASCII Authorization header crashed the request. Adds a catch-all, maps conflict to 409, backend unavailability to 503 and unexpected faults to 500, and documents the full response contract with the retry semantics each status implies, since senders key their behaviour off it. T07 - ingestion tests 2 -> 23, suite 15 -> 36. accepted_at is now UTC rather than local time, and naive timestamps are rejected instead of silently assumed. Remaining in WP-0004: T03 tenant/source binding, T04 redaction policy, T05 operator read surface, T06 production serving layer. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-10 14:30:20 +02:00
ON CONFLICT(event_id) DO NOTHING
RETURNING event_id
""",
(
event.event_id,
payload_hash,
datetime.now(timezone.utc).replace(microsecond=0).isoformat(),
str(details.get("correlation_id") or "") or None,
event.tenant,
json.dumps(event.as_record(), sort_keys=True),
link,
previous,
Route ingestion through the backend contract; fix error semantics AUDIT-WP-0004 T01, T02, T07. T01 - ingestion wrote to SQLite directly and never called the AuditBackend contract, so a 202 meant a row existed rather than that a backend with a declared retention policy had accepted the event. Adds IdempotentAuditBackend to the contract: duplicate detection lives inside the backend so custody and idempotency state share a transaction and cannot diverge. SQLiteAuditBackend implements it with WAL, synchronous=FULL and a busy timeout. Ingestion now refuses any backend declaring durable=False, so the development file backend cannot silently become the production sink. The atomicity claim was tested rather than asserted, and the first attempt failed: with a single shared connection, 16 racing submissions of one event told two callers they were first. Storage was correct but the response was not. Fixed with per-thread connections and BEGIN IMMEDIATE around the insert/read pair, and locked in by a test. T02 - storage errors previously escaped the handler with start_response never called, and the auth check sat outside the try block so a non-ASCII Authorization header crashed the request. Adds a catch-all, maps conflict to 409, backend unavailability to 503 and unexpected faults to 500, and documents the full response contract with the retry semantics each status implies, since senders key their behaviour off it. T07 - ingestion tests 2 -> 23, suite 15 -> 36. accepted_at is now UTC rather than local time, and naive timestamps are rejected instead of silently assumed. Remaining in WP-0004: T03 tenant/source binding, T04 redaction policy, T05 operator read surface, T06 production serving layer. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-10 14:30:20 +02:00
),
).fetchone()
existing = None
if inserted is None:
existing = db.execute(
"SELECT payload_hash FROM events WHERE event_id = ?", (event.event_id,)
).fetchone()
db.execute("COMMIT")
except sqlite3.Error as exc:
_rollback(db)
raise BackendUnavailableError(str(exc)) from exc
except BaseException:
_rollback(db)
raise
if inserted is not None:
return AcceptResult(duplicate=False, reference=reference)
if existing is None:
# The row vanished between the insert and the read inside one
# transaction, which should be impossible. Retryable rather than
# guessed at.
raise BackendUnavailableError("event disappeared during accept")
if existing[0] != payload_hash:
raise EventConflictError(
f"event_id {event.event_id} already held with a different payload"
)
return AcceptResult(duplicate=True, reference=reference)
Bind sender identities, add operator read surface, real serving layer AUDIT-WP-0004 T03, T05, T06. T03 - the receiver accepted whatever tenant and source a caller sent as long as it held the one shared token, despite WP-0003 recording tenant isolation as delivered. audit_core.senders binds each credential to the sources and tenants it may assert, driven by AUDIT_CORE_SENDERS rather than literals. Identities hold a list of tokens so rotation publishes the replacement alongside the incumbent and needs no delivery gap. Read is a separate privilege from write, so a sender credential cannot read the audit trail back. T05 - lookup by event id, lookup by correlation id, and a dead-letter view. Rejections are recorded rather than silently dropped. An event rejected for carrying secret-shaped material has its payload withheld: storing it would write that material into the audit store, which is what the rejection exists to prevent. Reason and payload hash are kept so it stays traceable. Replay is deliberately not built here. Idempotent replay is a property of the durable store and building it against SQLite would produce a second implementation to throw away; it lands with the Postgres backend in AUDIT-WP-0005-T01. T06 - serving moves to waitress with configurable threads and channel timeout, installed in the image via the serve extra. Without it the entrypoint falls back to a threaded wsgiref server with a socket timeout and graceful shutdown on SIGTERM, and logs a warning so a deployment cannot quietly land on the fallback. Metric counters deferred to WP-0005-T03 to be designed against the real scrape path. Tests 36 -> 46, covering cross-tenant and cross-source refusal, token rotation, read/write privilege separation, correlation lookup, and payload withholding on secret rejection. Remaining in WP-0004: T04 redaction policy, which needs a decision on whether a secret-shaped field is a rejection or a redaction. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-10 14:50:02 +02:00
# --- operator read surface (AUDIT-WP-0004-T05) --------------------------
def get(self, event_id: str) -> dict | None:
"""Return one stored event record, or None."""
row = self._query(
"SELECT record, accepted_at FROM events WHERE event_id = ?", (event_id,)
)
if not row:
return None
return {"accepted_at": row[0][1], **json.loads(row[0][0])}
def by_correlation(self, correlation_id: str, limit: int = 100) -> list[dict]:
"""Return every stored event carrying ``correlation_id``, oldest first."""
rows = self._query(
"SELECT record, accepted_at FROM events WHERE correlation_id = ? "
"ORDER BY accepted_at, event_id LIMIT ?",
(correlation_id, int(limit)),
)
return [{"accepted_at": at, **json.loads(rec)} for rec, at in rows]
def record_rejection(
self,
*,
event_id: str | None,
reason: str,
payload_hash: str,
sender: str | None = None,
payload: str | None = None,
) -> None:
"""Record a rejected event so it is visible to an operator.
A rejection is not a silent drop: the sender dead-letters the event and
somebody has to be able to see why. The payload is withheld when the
rejection reason implies it carries secret-shaped material.
"""
withheld = reason in WITHHOLD_PAYLOAD_REASONS
try:
self.db.execute(
"INSERT INTO dead_letters "
"(event_id, received_at, sender, reason, payload_hash, payload, payload_withheld) "
"VALUES (?, ?, ?, ?, ?, ?, ?)",
(
event_id,
datetime.now(timezone.utc).replace(microsecond=0).isoformat(),
sender,
reason,
payload_hash,
None if withheld else payload,
1 if withheld else 0,
),
)
except sqlite3.Error as exc:
raise BackendUnavailableError(str(exc)) from exc
def dead_letters(self, limit: int = 100) -> list[dict]:
"""Return recent rejections, newest first."""
rows = self._query(
"SELECT event_id, received_at, sender, reason, payload_hash, payload, "
"payload_withheld FROM dead_letters ORDER BY id DESC LIMIT ?",
(int(limit),),
)
return [
{
"event_id": r[0],
"received_at": r[1],
"sender": r[2],
"reason": r[3],
"payload_hash": r[4],
"payload": r[5],
"payload_withheld": bool(r[6]),
}
for r in rows
]
2026-08-10 16:02:22 +02:00
def count_secret_findings(
self, *, sender: str, source: str, action: str, outcome: str, findings
) -> None:
"""Increment the per-path counter for each finding.
Written durably rather than held in memory: these counters exist to
drive a fix in the sending service, and that work outlives a pod
restart.
"""
now = datetime.now(timezone.utc).replace(microsecond=0).isoformat()
try:
for finding in findings:
self.db.execute(
"""
INSERT INTO secret_findings
(sender, source, action, field_path, outcome, persisted,
occurrences, first_seen, last_seen)
VALUES (?, ?, ?, ?, ?, ?, 1, ?, ?)
ON CONFLICT(sender, source, action, field_path, outcome)
DO UPDATE SET occurrences = occurrences + 1, last_seen = excluded.last_seen
""",
(
sender, source, action, finding.path, outcome,
1 if finding.in_persisted_data else 0, now, now,
),
)
except sqlite3.Error as exc:
raise BackendUnavailableError(str(exc)) from exc
def secret_findings(self, limit: int = 100) -> list[dict]:
"""Return secret-shaped field counters, most frequent first."""
rows = self._query(
"SELECT sender, source, action, field_path, outcome, persisted, "
"occurrences, first_seen, last_seen FROM secret_findings "
"ORDER BY occurrences DESC, last_seen DESC LIMIT ?",
(int(limit),),
)
return [
{
"sender": r[0], "source": r[1], "action": r[2], "field_path": r[3],
"outcome": r[4], "persisted": bool(r[5]), "occurrences": r[6],
"first_seen": r[7], "last_seen": r[8],
}
for r in rows
]
Bind sender identities, add operator read surface, real serving layer AUDIT-WP-0004 T03, T05, T06. T03 - the receiver accepted whatever tenant and source a caller sent as long as it held the one shared token, despite WP-0003 recording tenant isolation as delivered. audit_core.senders binds each credential to the sources and tenants it may assert, driven by AUDIT_CORE_SENDERS rather than literals. Identities hold a list of tokens so rotation publishes the replacement alongside the incumbent and needs no delivery gap. Read is a separate privilege from write, so a sender credential cannot read the audit trail back. T05 - lookup by event id, lookup by correlation id, and a dead-letter view. Rejections are recorded rather than silently dropped. An event rejected for carrying secret-shaped material has its payload withheld: storing it would write that material into the audit store, which is what the rejection exists to prevent. Reason and payload hash are kept so it stays traceable. Replay is deliberately not built here. Idempotent replay is a property of the durable store and building it against SQLite would produce a second implementation to throw away; it lands with the Postgres backend in AUDIT-WP-0005-T01. T06 - serving moves to waitress with configurable threads and channel timeout, installed in the image via the serve extra. Without it the entrypoint falls back to a threaded wsgiref server with a socket timeout and graceful shutdown on SIGTERM, and logs a warning so a deployment cannot quietly land on the fallback. Metric counters deferred to WP-0005-T03 to be designed against the real scrape path. Tests 36 -> 46, covering cross-tenant and cross-source refusal, token rotation, read/write privilege separation, correlation lookup, and payload withholding on secret rejection. Remaining in WP-0004: T04 redaction policy, which needs a decision on whether a secret-shaped field is a rejection or a redaction. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-10 14:50:02 +02:00
def _query(self, sql: str, params: tuple) -> list:
try:
return self.db.execute(sql, params).fetchall()
except sqlite3.Error as exc:
raise BackendUnavailableError(str(exc)) from exc
def verify_chain(self, attestation: dict | None = None):
rows = self._query(
"SELECT event_id, payload_hash, chain_hash, chain_prev, accepted_at "
"FROM events ORDER BY accepted_at, event_id",
(),
)
return verify_rows(
[
ChainRow(
event_id=r[0],
payload_hash=r[1],
chain_hash=r[2] or "",
chain_prev=r[3] or "",
accepted_at=r[4],
)
for r in rows
],
attestation=attestation,
)
def attest_chain(self) -> dict:
return attestation_from_report(self.verify_chain())
def _ensure_chain_columns(self, db: sqlite3.Connection) -> None:
cols = {row[1] for row in db.execute("PRAGMA table_info(events)")}
if "chain_hash" not in cols:
db.execute("ALTER TABLE events ADD COLUMN chain_hash TEXT")
if "chain_prev" not in cols:
db.execute("ALTER TABLE events ADD COLUMN chain_prev TEXT")
missing = db.execute(
"SELECT event_id, payload_hash FROM events "
"WHERE chain_hash IS NULL OR chain_prev IS NULL "
"ORDER BY accepted_at, event_id"
).fetchall()
if not missing:
return
prev = GENESIS
# Recompute the whole chain so a partial backfill cannot fork.
for event_id, payload_hash in db.execute(
"SELECT event_id, payload_hash FROM events ORDER BY accepted_at, event_id"
):
link = chain_link(prev, payload_hash, event_id)
db.execute(
"UPDATE events SET chain_prev = ?, chain_hash = ? WHERE event_id = ?",
(prev, link, event_id),
)
prev = link
Route ingestion through the backend contract; fix error semantics AUDIT-WP-0004 T01, T02, T07. T01 - ingestion wrote to SQLite directly and never called the AuditBackend contract, so a 202 meant a row existed rather than that a backend with a declared retention policy had accepted the event. Adds IdempotentAuditBackend to the contract: duplicate detection lives inside the backend so custody and idempotency state share a transaction and cannot diverge. SQLiteAuditBackend implements it with WAL, synchronous=FULL and a busy timeout. Ingestion now refuses any backend declaring durable=False, so the development file backend cannot silently become the production sink. The atomicity claim was tested rather than asserted, and the first attempt failed: with a single shared connection, 16 racing submissions of one event told two callers they were first. Storage was correct but the response was not. Fixed with per-thread connections and BEGIN IMMEDIATE around the insert/read pair, and locked in by a test. T02 - storage errors previously escaped the handler with start_response never called, and the auth check sat outside the try block so a non-ASCII Authorization header crashed the request. Adds a catch-all, maps conflict to 409, backend unavailability to 503 and unexpected faults to 500, and documents the full response contract with the retry semantics each status implies, since senders key their behaviour off it. T07 - ingestion tests 2 -> 23, suite 15 -> 36. accepted_at is now UTC rather than local time, and naive timestamps are rejected instead of silently assumed. Remaining in WP-0004: T03 tenant/source binding, T04 redaction policy, T05 operator read surface, T06 production serving layer. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-10 14:30:20 +02:00
def health(self) -> None:
"""Raise :class:`BackendUnavailableError` if the store is unusable."""
try:
self.db.execute("SELECT 1 FROM events LIMIT 1").fetchone()
except sqlite3.Error as exc:
raise BackendUnavailableError(str(exc)) from exc
def close(self) -> None:
"""Close this thread's connection, if it has one."""
conn = getattr(self._local, "conn", None)
if conn is not None:
conn.close()
self._local.conn = None
def _rollback(db: sqlite3.Connection) -> None:
try:
db.execute("ROLLBACK")
except sqlite3.Error:
pass
def _record_hash(event: AuditEvent) -> str:
import hashlib
return hashlib.sha256(
json.dumps(event.as_record(), sort_keys=True).encode("utf-8")
).hexdigest()