Create every SQLite file 0600 (closes #255)

webhooker.db holds target configuration in plaintext — bearer tokens,
API keys, Slack webhook URLs — and the session encryption key, and it
was created 0644. The 0750 data directory was therefore the only
barrier, and a Docker bind mount supplies that directory at 0755,
which removes it: every local user on the host could read every stored
credential.

The mode is settled in OpenSQLite, the single open path all three
tiers share, so the main database, the per-webhook event databases and
the archive databases are covered in one place.

0644 comes from SQLite itself: robust_open substitutes
SQLITE_DEFAULT_FILE_PERMISSIONS whenever it is handed mode 0, and
findCreateFileMode yields 0 for a main database opened by URI with no
`modeof` parameter. A chmod after opening would leave a window in
which the credentials are on disk world-readable, so OpenSQLite
creates the file itself at 0600 before the driver sees the path.

That also settles the WAL sidecars, which carry the same rows and
would otherwise leave the fix worthless. SQLite derives both from the
main database file — `-wal` through findCreateFileMode, which stats
the path with the suffix stripped, and `-shm` in unixOpenSharedMemory
from an fstat of the open database descriptor — so a main file at 0600
produces sidecars at 0600. Verified by stat rather than by reading the
driver: a pre-change build leaves webhooker.db, -wal and -shm all 644;
this one leaves all three 600, in a 0755 bind mount, for all three
tiers, with delivery and restart working.

Existing files are chmodded on open, so a directory an earlier build
left 0644 — including a developer's scratch directory — is fixed
without any migration machinery.

This is not encryption at rest. An unattended process needs a key it
can read without a human, so the key lands beside the data and an
attacker who can read the database can read it too.

The data directory stays 0750: the group bit may matter to a
deployment, and with the files at 0600 the directory is no longer the
barrier.

README: the Docker section now states the DATA_DIR ownership
requirement where the bind mount is documented. A `-v` source path
Docker creates is root:root, and the container runs as UID 1000, so it
fails to start on the lock file; the chown that fixes it appeared only
under Restore.

README.md

@@ -71,8 +71,18 @@ make clean      # Remove bin/
 ### Configuration
 
 All configuration is via environment variables. For local development,
-you can place variables in a `.env` file in the project root (loaded
+you can place variables in a `.env` file in the process working
-automatically via `godotenv/autoload`).
+directory, read once at startup before anything else looks at the
+environment.
+
+The file is optional and having none is the normal case for a
+deployment. A file that is there but cannot be parsed aborts startup
+with a message naming it, because a single malformed line makes none
+of the file apply: every variable in it silently reverts to its
+default, which is exactly the failure [Invalid values abort
+startup](#invalid-values-abort-startup) exists to prevent, for all of
+them at once. A variable already present in the real environment wins
+over the file's value for the same name.
 
 The environment is selected by setting `WEBHOOKER_ENVIRONMENT` to `dev`
 or `prod` (default: `dev`). The setting controls exactly one behavior:
@@ -125,7 +135,7 @@ TTY detection, and security headers are always applied.
 | `MAINTENANCE_MODE`      | Report `maintenanceMode: true` in the healthcheck JSON. It does not change how any request is served — no maintenance page exists | `false`  |
 | `METRICS_USERNAME`      | Basic auth username for `/metrics`. Must be set together with `METRICS_PASSWORD`; one without the other fails startup | `""`     |
 | `METRICS_PASSWORD`      | Basic auth password for `/metrics`. Must be set together with `METRICS_USERNAME`; one without the other fails startup | `""`     |
-| `SENTRY_DSN`            | Sentry error reporting DSN          | `""`     |
+| `SENTRY_DSN`            | Sentry error reporting DSN. Unset leaves error reporting off; a value the Sentry SDK cannot parse fails startup rather than serving with reporting silently off | `""`     |
 | `RETENTION_SWEEP_INTERVAL` | How often the retention reaper and archive sweeper run (Go duration, must be positive) | `1h` |
 | `SESSION_IDLE_TIMEOUT`  | Idle session timeout (Go duration)  | `24h`    |
 | `RECEIVER_RATE_LIMIT`   | Receiver requests/minute per IP per entrypoint (10x that per IP across the route) | `120` |
@@ -464,10 +474,20 @@ startup), every entry in `TRUSTED_PROXIES` and
 `ALLOWED_EGRESS_CIDRS` must be a CIDR block or a bare IP address, and
 `BIND_ADDRESS` must be an IP address literal — `localhost`,
 `127.0.0.1:8080` and `10.0.0.0/8` are each rejected rather than
-resolved, split, or narrowed to something they do not say.
+resolved, split, or narrowed to something they do not say — and
+`SENTRY_DSN` must parse as a Sentry DSN.
 `SESSION_IDLE_TIMEOUT` is the exception: a
 non-positive value there means idle expiry is disabled, not invalid.
 
+`SENTRY_DSN` is checked with the Sentry SDK's own parser, the same call
+the SDK makes on the DSN it is later handed, so what configuration
+accepts is exactly what will initialise. A typo in it is the one
+configuration mistake nothing downstream can ever notice — the variable
+is still set, so every later signal reports error reporting as on while
+no report is being sent — which is why it aborts rather than starting
+with reporting off. Leaving it unset is not a mistake and not affected:
+error reporting is simply off and startup is normal.
+
 Boolean variables (`DEBUG`, `MAINTENANCE_MODE`) accept exactly the
 spellings Go's `strconv.ParseBool` accepts — `1`, `t`, `T`, `TRUE`,
 `true`, `True`, `0`, `f`, `F`, `FALSE`, `false`, `False` — and nothing
@@ -666,6 +686,44 @@ databases written by `database` targets (`archive-{uuid}.db`). Mount
 this as a persistent volume to preserve data across container
 restarts.
 
+**The bind-mounted directory must be owned by UID 1000, or the
+container does not start.** Docker creates a `-v` source path that
+does not exist yet as `root:root`, and the process runs as UID 1000,
+so it cannot take its `DATA_DIR` lock:
+
+```
+webhooker: locking data directory /var/lib/webhooker: open
+/var/lib/webhooker/webhooker.lock: permission denied
+```
+
+It exits non-zero at that point, before opening any database. Create
+the directory ahead of the first `docker run`:
+
+```bash
+mkdir -p /path/to/data
+chown 1000:1000 /path/to/data
+chmod 750 /path/to/data
+```
+
+The same `chown` is what a restore needs — see step 4 of
+[Restore](#restore). A **named volume** does not have this problem:
+Docker copies the image's ownership onto a volume it initializes, and
+the image creates `/var/lib/webhooker` owned by `webhooker`.
+
+**The file modes are not yours to set, and do not depend on the
+directory.** `webhooker.db` holds target configuration in plaintext —
+bearer tokens, API keys, Slack webhook URLs — along with the session
+encryption key, so webhooker creates every SQLite file it owns `0600`:
+each database and both of its `-wal` and `-shm` sidecars, across all
+three tiers. Files an earlier build left `0644` are tightened when
+they are opened. A `DATA_DIR` webhooker creates itself is `0750`, but
+a bind mount supplies its own directory and Docker's default for one
+it creates is `0755`; the `0600` files hold there regardless. The
+`chmod 750` above is defence in depth — it stops other local users
+listing the directory and learning your webhook UUIDs from the
+`events-{uuid}.db` filenames — not the barrier protecting the
+credentials.
+
 ## Deployment behind a reverse proxy
 
 webhooker terminates no TLS of its own. It serves plaintext HTTP and
@@ -1622,6 +1680,11 @@ webhooker uses **separate SQLite database files**: a main application
 database for configuration data and per-webhook databases for event
 storage. All database files live in the `DATA_DIR` directory.
 
+Every one of them is created `0600`, and so is each `-wal` and `-shm`
+sidecar. See
+[Running with Docker](#running-with-docker) for what that does and
+does not protect.
+
 **Main Application Database** (`{DATA_DIR}/webhooker.db`) — stores
 configuration and application state:
 

cmd/webhooker/dotenv_test.go

@@ -0,0 +1,107 @@
+package main
+
+import (
+	"bytes"
+	"os"
+	"path/filepath"
+	"strings"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+	"sneak.berlin/go/webhooker/internal/config"
+)
+
+// dotEnvKey is a throwaway variable name these tests write and read,
+// so they cannot disturb real configuration.
+const dotEnvKey = "WEBHOOKER_TEST_DISPATCH_VALUE"
+
+// writeDotEnvInWorkingDir puts contents in a .env file in a fresh
+// temporary directory and moves the process there.
+//
+// The callers are deliberately not parallel and must stay that way:
+// t.Chdir moves the whole process. Go releases parallel tests only
+// after every sequential test in the package has finished, so nothing
+// else runs while these do.
+func writeDotEnvInWorkingDir(t *testing.T, contents string) {
+	t.Helper()
+
+	dir := t.TempDir()
+	require.NoError(t, os.WriteFile(
+		filepath.Join(dir, config.DotEnvPath),
+		[]byte(contents), 0o600,
+	))
+	t.Chdir(dir)
+}
+
+// TestDispatch_MalformedDotEnvRefuses pins the second half of the
+// defect. godotenv applies nothing at all when a file will not parse,
+// so one mistyped line used to revert every variable in it to its
+// default and start the server anyway, with no log line naming the
+// file. The refusal has to arrive before any subcommand runs, which
+// is why `help` — the one subcommand that touches nothing — is still
+// refused here.
+//
+//nolint:paralleltest // t.Chdir moves the whole process.
+func TestDispatch_MalformedDotEnvRefuses(t *testing.T) {
+	writeDotEnvInWorkingDir(t, "PORT 19615\n")
+
+	var stdout, stderr bytes.Buffer
+
+	code := dispatch(
+		[]string{helpCommand}, strings.NewReader(""), &stdout, &stderr,
+	)
+
+	require.Equal(t, 1, code, "a broken .env must exit non-zero")
+	assert.Contains(
+		t, stderr.String(), config.DotEnvPath,
+		"the refusal must name the file",
+	)
+	assert.Empty(
+		t, stdout.String(),
+		"the subcommand must not have run",
+	)
+}
+
+// TestDispatch_LoadsDotEnvBeforeSubcommands pins the ordering the
+// godotenv/autoload import used to provide for free. It ran in an
+// init(), so .env was in the environment before anything read it —
+// including config.DataDir, which both the DATA_DIR lock and resetpw
+// call outside the fx graph. Loading any later would let a .env that
+// sets DATA_DIR lock one directory while the config opened databases
+// in another.
+func TestDispatch_LoadsDotEnvBeforeSubcommands(t *testing.T) {
+	t.Setenv(dotEnvKey, "placeholder")
+	require.NoError(t, os.Unsetenv(dotEnvKey))
+
+	writeDotEnvInWorkingDir(t, dotEnvKey+"=from-dot-env\n")
+
+	var stdout, stderr bytes.Buffer
+
+	code := dispatch(
+		[]string{helpCommand}, strings.NewReader(""), &stdout, &stderr,
+	)
+
+	require.Equal(t, 0, code)
+	assert.Equal(
+		t, "from-dot-env", os.Getenv(dotEnvKey),
+		"the file must be applied before the subcommand runs",
+	)
+}
+
+// TestDispatch_MissingDotEnvIsFine pins the case most deployments are
+// in: no .env at all, which must stay a normal start.
+//
+//nolint:paralleltest // t.Chdir moves the whole process.
+func TestDispatch_MissingDotEnvIsFine(t *testing.T) {
+	t.Chdir(t.TempDir())
+
+	var stdout, stderr bytes.Buffer
+
+	code := dispatch(
+		[]string{helpCommand}, strings.NewReader(""), &stdout, &stderr,
+	)
+
+	require.Equal(t, 0, code)
+	assert.Empty(t, stderr.String())
+}

cmd/webhooker/main.go

@@ -54,6 +54,11 @@ const stopTimeout = 5 * time.Second
 // caller can tell "called wrong" from "declined".
 const exitUsage = 2
 
+// helpCommand is the subcommand that prints usage. The flag spellings
+// beside it in the switch are aliases; this is the name the usage text
+// documents and the one tests invoke.
+const helpCommand = "help"
+
 // Build-time variables set via -ldflags.
 //
 //nolint:gochecknoglobals // Build-time variables injected by the linker.
@@ -75,11 +80,27 @@ func main() {
 // every existing deployment invoke; that path is unchanged, including
 // where the DATA_DIR lock is taken relative to building the fx graph
 // and how fx propagates a non-zero exit itself.
+//
+// The optional .env file is read here, before any subcommand and so
+// before anything reads the environment — config.DataDir, which both
+// the DATA_DIR lock and resetpw call outside the fx graph, above all.
+// It used to be read from an init() in internal/config, which put it
+// earlier still but threw the error away: a single malformed line
+// applied none of the file and said nothing about it. A file that is
+// not there stays fine, since .env is optional and most deployments
+// do not have one.
 func dispatch(
 	args []string,
 	stdin io.Reader,
 	stdout, stderr io.Writer,
 ) int {
+	err := config.LoadDotEnv()
+	if err != nil {
+		_, _ = fmt.Fprintf(stderr, "%s: %v\n", appname, err)
+
+		return 1
+	}
+
 	if len(args) == 0 {
 		return run(stderr)
 	}
@@ -87,7 +108,7 @@ func dispatch(
 	switch args[0] {
 	case resetpw.Name:
 		return resetpw.Run(args[1:], stdin, stdout, stderr)
-	case "help", "-h", "-help", "--help":
+	case helpCommand, "-h", "-help", "--help":
 		usage(stdout)
 
 		return 0

cmd/webhooker/main_test.go

@@ -121,7 +121,7 @@ func TestDispatch_Help(t *testing.T) {
 	var stdout, stderr bytes.Buffer
 
 	code := dispatch(
-		[]string{"help"}, strings.NewReader(""), &stdout, &stderr,
+		[]string{helpCommand}, strings.NewReader(""), &stdout, &stderr,
 	)
 
 	require.Equal(t, 0, code)

internal/config/config.go

@@ -4,6 +4,7 @@ package config
 import (
 	"errors"
 	"fmt"
+	"io/fs"
 	"log/slog"
 	"net/netip"
 	"os"
@@ -11,13 +12,11 @@ import (
 	"strings"
 	"time"
 
+	"github.com/getsentry/sentry-go"
+	"github.com/joho/godotenv"
 	"go.uber.org/fx"
 	"sneak.berlin/go/webhooker/internal/globals"
 	"sneak.berlin/go/webhooker/internal/logger"
-
-	// Populates the environment from a ./.env file automatically for
-	// development configuration. Kept in one place only (here).
-	_ "github.com/joho/godotenv/autoload"
 )
 
 const (
@@ -84,6 +83,12 @@ const (
 	// IPv6 prefix spends on the ::ffff:0:0/96 wrapper, so a /104
 	// covers the same addresses as an IPv4 /8.
 	mappedV4Offset = 96
+
+	// DotEnvPath is the optional file of KEY=value lines read into the
+	// environment at startup, relative to the process working
+	// directory. Exported so that documentation and tests name the
+	// same path the loader opens.
+	DotEnvPath = ".env"
 )
 
 // ErrInvalidEnvironment is returned when WEBHOOKER_ENVIRONMENT
@@ -107,6 +112,15 @@ var ErrInvalidCIDR = errors.New("invalid CIDR")
 // something that is not an IP address literal.
 var ErrInvalidBindAddress = errors.New("invalid bind address")
 
+// ErrInvalidSentryDSN is returned when SENTRY_DSN is set to something
+// the Sentry SDK cannot parse as a DSN.
+var ErrInvalidSentryDSN = errors.New("invalid Sentry DSN")
+
+// ErrDotEnvUnreadable is returned when the optional .env file exists
+// but cannot be read or parsed. A file that is not there is not an
+// error; a file that is there and broken is.
+var ErrDotEnvUnreadable = errors.New("unreadable .env file")
+
 // ErrIncompleteMetricsAuth is returned when exactly one of
 // METRICS_USERNAME and METRICS_PASSWORD carries a value. Neither
 // fallback is acceptable: serving /metrics on the username alone
@@ -212,12 +226,62 @@ func (c *Config) MetricsAuthEnabled() bool {
 	return c.MetricsUsername != "" && c.MetricsPassword != ""
 }
 
+// SentryEnabled reports whether error reporting is shipped to Sentry.
+// It is the only answer to that question in the codebase: the SDK
+// initialisation, the sentryhttp middleware registration and the
+// startup log's sentryEnabled field all read this one method, so the
+// log cannot report reporting as on while nothing is sending.
+//
+// A non-empty DSN is enough because loadFromEnv already parsed it with
+// the SDK's own parser and refused to build a Config around one the
+// SDK would reject, and because initialising the SDK with a DSN that
+// parsed and failed anyway aborts the process rather than leaving this
+// true and the client absent.
+func (c *Config) SentryEnabled() bool {
+	return c.SentryDSN != ""
+}
+
 // envString returns the value of the named environment variable,
 // or an empty string if not set.
 func envString(key string) string {
 	return os.Getenv(key)
 }
 
+// LoadDotEnv reads DotEnvPath into the environment when that file is
+// present, and reports a file that is present but broken.
+//
+// It has to run before anything reads the environment, so that every
+// reader agrees on what the environment holds — the DATA_DIR lock
+// taken before the fx graph exists as much as loadFromEnv itself. A
+// variable already set in the real environment wins: godotenv never
+// overwrites one.
+//
+// A missing file is not an error. It is a development convenience and
+// most deployments set the environment directly.
+//
+// Any other failure is. godotenv parses the whole file before setting
+// anything, so a single malformed line applies none of it: every
+// variable in the file silently reverts to its default, which defeats
+// the fail-loud guarantee for all of them at once.
+func LoadDotEnv() error {
+	return loadDotEnvFile(DotEnvPath)
+}
+
+// loadDotEnvFile is LoadDotEnv over a named file, so tests can point
+// at a temporary one instead of the process working directory.
+func loadDotEnvFile(path string) error {
+	err := godotenv.Load(path)
+	if err == nil || errors.Is(err, fs.ErrNotExist) {
+		return nil
+	}
+
+	return fmt.Errorf(
+		"%w: %s: %w; nothing in it was applied, so fix the file or "+
+			"remove it",
+		ErrDotEnvUnreadable, path, err,
+	)
+}
+
 // DataDir resolves DATA_DIR, applying DefaultDataDir when it is unset
 // or empty. It is exported so that entry points which must act on the
 // data directory before the fx graph exists — taking the exclusive
@@ -462,6 +526,41 @@ func envBindAddress(key, defaultValue string) (string, error) {
 	return addr.String(), nil
 }
 
+// envSentryDSN returns the value of the named environment variable
+// checked as a Sentry DSN. An unset (or empty, or whitespace-only)
+// value yields "", which means error reporting stays off — the common
+// case, and a normal start.
+//
+// A set value is parsed with sentry.NewDsn, which is the call
+// sentry.Init makes on the DSN it is handed, so what passes here is
+// exactly what the SDK will accept later and the two cannot disagree.
+// Reproducing the check by hand instead would cost this package its
+// dependency on the SDK — already a module dependency, already linked
+// into the binary — in exchange for a second definition of "valid DSN"
+// free to drift from the one that decides.
+//
+// A set value that does not parse is a hard error naming the key, so
+// startup fails loudly. Losing error reporting is the failure this
+// variable exists to prevent, and a typo in a DSN is silent forever:
+// nothing later in the process can notice that reports are going
+// nowhere. The bad value is quoted because it is a URL to a public
+// endpoint carrying a public key, not a secret.
+func envSentryDSN(key string) (string, error) {
+	v := strings.TrimSpace(os.Getenv(key))
+	if v == "" {
+		return "", nil
+	}
+
+	_, err := sentry.NewDsn(v)
+	if err != nil {
+		return "", fmt.Errorf(
+			"%w: %s: %q: %w", ErrInvalidSentryDSN, key, v, err,
+		)
+	}
+
+	return v, nil
+}
+
 // resolveMetricsAuth reads the /metrics basic-auth credentials and
 // rejects a half-set pair, naming both variables either way. The
 // error carries neither value: the password is a secret.
@@ -594,6 +693,11 @@ func loadFromEnv() (*Config, error) {
 		return nil, err
 	}
 
+	sentryDSN, err := envSentryDSN("SENTRY_DSN")
+	if err != nil {
+		return nil, err
+	}
+
 	return &Config{
 		DataDir:                DataDir(),
 		Debug:                  debug,
@@ -603,7 +707,7 @@ func loadFromEnv() (*Config, error) {
 		MetricsPassword:        metricsPassword,
 		Port:                   port,
 		BindAddress:            bindAddress,
-		SentryDSN:              envString("SENTRY_DSN"),
+		SentryDSN:              sentryDSN,
 		RetentionSweepInterval: retentionSweepInterval,
 		SessionIdleTimeout:     sessionIdleTimeout,
 		ReceiverRateLimit:      receiverRateLimit,
@@ -738,7 +842,7 @@ func New(lc fx.Lifecycle, params ConfigParams) (*Config, error) {
 		"receiverRateLimit", s.ReceiverRateLimit,
 		"trustedProxies", len(s.TrustedProxies),
 		"allowedEgressCIDRs", len(s.AllowedEgressCIDRs),
-		"hasSentryDSN", s.SentryDSN != "",
+		"sentryEnabled", s.SentryEnabled(),
 		"hasMetricsAuth", s.MetricsAuthEnabled(),
 	)
 

internal/config/dotenv_test.go

@@ -0,0 +1,158 @@
+package config_test
+
+import (
+	"os"
+	"path/filepath"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+	"sneak.berlin/go/webhooker/internal/config"
+)
+
+// dotEnvKey is a throwaway variable name the .env tests write and
+// read, so they cannot disturb real configuration.
+const dotEnvKey = "WEBHOOKER_TEST_DOTENV_VALUE"
+
+// malformedDotEnv is a file godotenv cannot parse. The first line is
+// the realistic typo — a space where the `=` belongs — and the rest
+// make sure nothing downstream treats the file as salvageable line by
+// line.
+const malformedDotEnv = "PORT 19615\n" +
+	"this is not = valid ! syntax\n" +
+	"\"unclosed\n"
+
+// unsetDotEnvKey makes dotEnvKey genuinely absent for the duration of
+// the test and restores it afterwards. t.Setenv registers the restore;
+// the Unsetenv that follows is what the test actually needs, because a
+// variable set to the empty string is still present in os.Environ and
+// godotenv would refuse to overwrite it.
+func unsetDotEnvKey(t *testing.T) {
+	t.Helper()
+	t.Setenv(dotEnvKey, "placeholder")
+	require.NoError(t, os.Unsetenv(dotEnvKey))
+}
+
+// writeDotEnv writes contents to a .env file in a fresh temporary
+// directory and returns its path.
+func writeDotEnv(t *testing.T, contents string) string {
+	t.Helper()
+
+	path := filepath.Join(t.TempDir(), config.DotEnvPath)
+	require.NoError(t, os.WriteFile(path, []byte(contents), 0o600))
+
+	return path
+}
+
+// TestLoadDotEnv_MissingFileIsFine pins the case most deployments are
+// in. The file is optional: it is a development convenience, and a
+// deployment that configures the environment directly must start
+// normally rather than be refused for a file it was never meant to
+// have.
+//
+//nolint:paralleltest // unsetDotEnvKey uses t.Setenv.
+func TestLoadDotEnv_MissingFileIsFine(t *testing.T) {
+	unsetDotEnvKey(t)
+
+	absent := filepath.Join(t.TempDir(), config.DotEnvPath)
+	require.NoError(t, config.LoadDotEnvFileForTest(absent))
+
+	_, present := os.LookupEnv(dotEnvKey)
+	assert.False(t, present, "nothing may be set from an absent file")
+}
+
+// TestLoadDotEnv_AppliesValues pins that a well-formed file still
+// reaches the environment, which is the whole reason the file is read
+// at all.
+//
+//nolint:paralleltest // unsetDotEnvKey uses t.Setenv.
+func TestLoadDotEnv_AppliesValues(t *testing.T) {
+	unsetDotEnvKey(t)
+
+	path := writeDotEnv(t, "# a comment\n"+dotEnvKey+"=from-dot-env\n")
+
+	require.NoError(t, config.LoadDotEnvFileForTest(path))
+	assert.Equal(t, "from-dot-env", os.Getenv(dotEnvKey))
+}
+
+// TestLoadDotEnv_RealEnvironmentWins pins that the file cannot
+// override a variable the process was actually started with. A
+// deployment that sets DATA_DIR in its unit file must not have it
+// silently replaced by a stale .env left in the working directory.
+func TestLoadDotEnv_RealEnvironmentWins(t *testing.T) {
+	t.Setenv(dotEnvKey, "from-environment")
+
+	path := writeDotEnv(t, dotEnvKey+"=from-dot-env\n")
+
+	require.NoError(t, config.LoadDotEnvFileForTest(path))
+	assert.Equal(t, "from-environment", os.Getenv(dotEnvKey))
+}
+
+// TestLoadDotEnv_MalformedFileAborts is the defect this fixes. One bad
+// line makes godotenv apply none of the file, so every variable in it
+// reverts to its default; the process used to start that way with no
+// log line naming the file at all.
+//
+//nolint:paralleltest // unsetDotEnvKey uses t.Setenv.
+func TestLoadDotEnv_MalformedFileAborts(t *testing.T) {
+	unsetDotEnvKey(t)
+
+	path := writeDotEnv(
+		t, malformedDotEnv+dotEnvKey+"=from-dot-env\n",
+	)
+
+	err := config.LoadDotEnvFileForTest(path)
+
+	require.Error(t, err)
+	require.ErrorIs(t, err, config.ErrDotEnvUnreadable)
+	assert.Contains(
+		t, err.Error(), config.DotEnvPath,
+		"the failure must name the file it could not read",
+	)
+
+	_, present := os.LookupEnv(dotEnvKey)
+	assert.False(
+		t, present,
+		"a rejected file must apply nothing, not part of itself",
+	)
+}
+
+// TestLoadDotEnv_UnreadableFileAborts pins that only absence is
+// tolerated. A .env that exists but cannot be read is a file the
+// operator meant to be applied, so it fails like a malformed one
+// rather than being treated as though it were not there.
+func TestLoadDotEnv_UnreadableFileAborts(t *testing.T) {
+	t.Parallel()
+
+	// A directory in the file's place: open succeeds and the read
+	// fails, which no umask or root-ness can turn back into success
+	// the way a chmod could.
+	path := filepath.Join(t.TempDir(), config.DotEnvPath)
+	require.NoError(t, os.Mkdir(path, 0o750))
+
+	err := config.LoadDotEnvFileForTest(path)
+
+	require.Error(t, err)
+	require.ErrorIs(t, err, config.ErrDotEnvUnreadable)
+}
+
+// TestLoadDotEnv_ReadsTheWorkingDirectory pins the path LoadDotEnv
+// itself opens, which the tests above bypass. It is relative to the
+// process working directory, as it was under godotenv/autoload and as
+// the README documents.
+//
+//nolint:paralleltest // t.Chdir moves the whole process.
+func TestLoadDotEnv_ReadsTheWorkingDirectory(t *testing.T) {
+	unsetDotEnvKey(t)
+
+	dir := t.TempDir()
+	require.NoError(t, os.WriteFile(
+		filepath.Join(dir, config.DotEnvPath),
+		[]byte(dotEnvKey+"=from-working-directory\n"),
+		0o600,
+	))
+	t.Chdir(dir)
+
+	require.NoError(t, config.LoadDotEnv())
+	assert.Equal(t, "from-working-directory", os.Getenv(dotEnvKey))
+}

internal/config/env_test.go

@@ -518,8 +518,20 @@ type badEnvValueCase struct {
 }
 
 // badEnvValueCases is the config.New table, kept out of the test body
-// so the test itself stays readable.
+// so the test itself stays readable. It is assembled from per-variable
+// groups because one literal covering every variable outgrew the
+// function-length budget.
 func badEnvValueCases() []badEnvValueCase {
+	cases := listenerEnvValueCases()
+	cases = append(cases, flagEnvValueCases()...)
+	cases = append(cases, sentryEnvValueCases()...)
+
+	return cases
+}
+
+// listenerEnvValueCases covers the two variables that describe the
+// HTTP listener.
+func listenerEnvValueCases() []badEnvValueCase {
 	return []badEnvValueCase{
 		{
 			name:  "valid PORT is used",
@@ -542,27 +554,6 @@ func badEnvValueCases() []badEnvValueCase {
 			value:       "70000",
 			expectError: true,
 		},
-		{
-			name:  "valid DEBUG is used",
-			key:   envKeyDebug,
-			value: "true",
-			check: func(t *testing.T, cfg *config.Config) {
-				t.Helper()
-				assert.True(t, cfg.Debug)
-			},
-		},
-		{
-			name:        "unparseable DEBUG aborts startup",
-			key:         envKeyDebug,
-			value:       "ture",
-			expectError: true,
-		},
-		{
-			name:        "unparseable MAINTENANCE_MODE aborts startup",
-			key:         envKeyMaintenanceMode,
-			value:       "sometimes",
-			expectError: true,
-		},
 		{
 			name:  "valid BIND_ADDRESS is used",
 			key:   envKeyBindAddress,
@@ -595,6 +586,69 @@ func badEnvValueCases() []badEnvValueCase {
 	}
 }
 
+// flagEnvValueCases covers the boolean variables.
+func flagEnvValueCases() []badEnvValueCase {
+	return []badEnvValueCase{
+		{
+			name:  "valid DEBUG is used",
+			key:   envKeyDebug,
+			value: "true",
+			check: func(t *testing.T, cfg *config.Config) {
+				t.Helper()
+				assert.True(t, cfg.Debug)
+			},
+		},
+		{
+			name:        "unparseable DEBUG aborts startup",
+			key:         envKeyDebug,
+			value:       "ture",
+			expectError: true,
+		},
+		{
+			name:        "unparseable MAINTENANCE_MODE aborts startup",
+			key:         envKeyMaintenanceMode,
+			value:       "sometimes",
+			expectError: true,
+		},
+	}
+}
+
+// sentryEnvValueCases covers SENTRY_DSN. The three rejected values are
+// the ones measured on the defect: each initialised the SDK with an
+// error and left the process serving with error reporting off.
+func sentryEnvValueCases() []badEnvValueCase {
+	return []badEnvValueCase{
+		{
+			name:  "valid SENTRY_DSN is used",
+			key:   envKeySentryDSN,
+			value: validSentryDSN,
+			check: func(t *testing.T, cfg *config.Config) {
+				t.Helper()
+				assert.Equal(t, validSentryDSN, cfg.SentryDSN)
+				assert.True(t, cfg.SentryEnabled())
+			},
+		},
+		{
+			name:        "unparseable SENTRY_DSN aborts startup",
+			key:         envKeySentryDSN,
+			value:       "not-a-dsn",
+			expectError: true,
+		},
+		{
+			name:        "SENTRY_DSN that is not a URL aborts startup",
+			key:         envKeySentryDSN,
+			value:       "%%%",
+			expectError: true,
+		},
+		{
+			name:        "keyless SENTRY_DSN aborts startup",
+			key:         envKeySentryDSN,
+			value:       "https://example.invalid/1",
+			expectError: true,
+		},
+	}
+}
+
 // TestNewUsesDefaultsWhenUnset proves the fail-loud behaviour did not
 // break the legitimate unset case: absent variables still get their
 // documented defaults.
@@ -603,7 +657,7 @@ func TestNewUsesDefaultsWhenUnset(t *testing.T) {
 
 	for _, key := range []string{
 		envKeyPort, envKeyDebug, envKeyMaintenanceMode,
-		envKeyBindAddress,
+		envKeyBindAddress, envKeySentryDSN,
 	} {
 		require.NoError(t, os.Unsetenv(key))
 	}
@@ -626,4 +680,9 @@ func TestNewUsesDefaultsWhenUnset(t *testing.T) {
 		t, config.DefaultBindAddressForTest, cfg.BindAddress,
 	)
 	assert.Equal(t, bindAddressDefault, cfg.BindAddress)
+
+	// An absent SENTRY_DSN is the common case and must stay a normal
+	// start with error reporting off, not a refusal.
+	assert.Empty(t, cfg.SentryDSN)
+	assert.False(t, cfg.SentryEnabled())
 }

internal/config/export_test.go

@@ -51,6 +51,18 @@ func EnvPortForTest(key string, defaultValue int) (int, error) {
 	return envPort(key, defaultValue)
 }
 
+// EnvSentryDSNForTest exposes envSentryDSN.
+func EnvSentryDSNForTest(key string) (string, error) {
+	return envSentryDSN(key)
+}
+
+// LoadDotEnvFileForTest exposes the loader LoadDotEnv runs, over a
+// caller-named file rather than the process working directory, so
+// each .env state can be covered without moving the test process.
+func LoadDotEnvFileForTest(path string) error {
+	return loadDotEnvFile(path)
+}
+
 // EnvBindAddressForTest exposes envBindAddress.
 func EnvBindAddressForTest(key, defaultValue string) (string, error) {
 	return envBindAddress(key, defaultValue)

internal/config/sentry_test.go

@@ -0,0 +1,141 @@
+package config_test
+
+import (
+	"os"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+	"sneak.berlin/go/webhooker/internal/config"
+)
+
+// envKeySentryDSN is the variable envSentryDSN reads in production.
+const envKeySentryDSN = "SENTRY_DSN"
+
+// validSentryDSN is a syntactically complete DSN. The host is under
+// .invalid (RFC 2606), so nothing a test builds around it can reach a
+// real Sentry installation.
+const validSentryDSN = "https://abc123@sentry.invalid/42"
+
+// envSentryDSNCase is one row of the envSentryDSN table.
+type envSentryDSNCase struct {
+	name        string
+	set         bool
+	value       string
+	expectError bool
+	expected    string
+}
+
+// envSentryDSNCases is the envSentryDSN table. The three invalid
+// values are the ones measured on the defect: each initialised the SDK
+// with an error and left the process serving with reporting off.
+func envSentryDSNCases() []envSentryDSNCase {
+	return []envSentryDSNCase{
+		{
+			name:     "unset means reporting off",
+			expected: "",
+		},
+		{
+			name:     "empty means reporting off",
+			set:      true,
+			value:    "",
+			expected: "",
+		},
+		{
+			name:     "whitespace means reporting off",
+			set:      true,
+			value:    "   ",
+			expected: "",
+		},
+		{
+			name:     "a valid DSN is kept",
+			set:      true,
+			value:    validSentryDSN,
+			expected: validSentryDSN,
+		},
+		{
+			name:     "surrounding whitespace is trimmed",
+			set:      true,
+			value:    "  " + validSentryDSN + "\t",
+			expected: validSentryDSN,
+		},
+		{
+			name:        "a value that is not a URL is rejected",
+			set:         true,
+			value:       "not-a-dsn",
+			expectError: true,
+		},
+		{
+			name:        "an unparseable URL is rejected",
+			set:         true,
+			value:       "%%%",
+			expectError: true,
+		},
+		{
+			name:        "a DSN without a public key is rejected",
+			set:         true,
+			value:       "https://example.invalid/1",
+			expectError: true,
+		},
+		{
+			name:        "a DSN without a project id is rejected",
+			set:         true,
+			value:       "https://abc123@sentry.invalid/",
+			expectError: true,
+		},
+		{
+			name:        "a non-HTTP scheme is rejected",
+			set:         true,
+			value:       "ftp://abc123@sentry.invalid/42",
+			expectError: true,
+		},
+	}
+}
+
+// TestEnvSentryDSN covers the helper directly. What it pins beyond the
+// value is the failure shape: a set-but-unparseable DSN names the
+// variable and the value, exactly as the other fail-loud helpers do,
+// so an operator reads the fix off the message.
+func TestEnvSentryDSN(t *testing.T) {
+	for _, tt := range envSentryDSNCases() {
+		t.Run(tt.name, func(t *testing.T) {
+			// Cannot use t.Parallel() here because t.Setenv
+			// is incompatible with parallel subtests.
+			if tt.set {
+				t.Setenv(envKeySentryDSN, tt.value)
+			} else {
+				require.NoError(t, os.Unsetenv(envKeySentryDSN))
+			}
+
+			got, err := config.EnvSentryDSNForTest(envKeySentryDSN)
+
+			if tt.expectError {
+				require.Error(t, err)
+				require.ErrorIs(t, err, config.ErrInvalidSentryDSN)
+				assert.Contains(t, err.Error(), envKeySentryDSN)
+				assert.Contains(t, err.Error(), tt.value)
+				assert.Empty(t, got)
+
+				return
+			}
+
+			require.NoError(t, err)
+			assert.Equal(t, tt.expected, got)
+		})
+	}
+}
+
+// TestSentryEnabled_TracksTheDSN pins that the one method answering
+// "is anything being reported" agrees with the DSN in every state. The
+// startup log, the SDK initialisation and the sentryhttp middleware
+// all read it, so a log field cannot report reporting as on while
+// nothing is sending.
+func TestSentryEnabled_TracksTheDSN(t *testing.T) {
+	t.Parallel()
+
+	assert.False(t, (&config.Config{}).SentryEnabled())
+	assert.True(
+		t,
+		(&config.Config{SentryDSN: validSentryDSN}).SentryEnabled(),
+	)
+}

internal/database/sqlite_mode_test.go

@@ -0,0 +1,240 @@
+package database_test
+
+import (
+	"context"
+	"io/fs"
+	"net/http"
+	"os"
+	"path/filepath"
+	"testing"
+
+	"github.com/google/uuid"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+	"go.uber.org/fx/fxtest"
+	"sneak.berlin/go/webhooker/internal/config"
+	"sneak.berlin/go/webhooker/internal/database"
+	"sneak.berlin/go/webhooker/internal/globals"
+	"sneak.berlin/go/webhooker/internal/logger"
+)
+
+// ownerOnly is the mode every SQLite file the service owns must have.
+// Spelled out rather than referencing database.SQLiteFilePerm so the
+// test fails if the constant itself is loosened.
+const ownerOnly fs.FileMode = 0o600
+
+// requireOwnerOnly asserts that path exists and is readable and
+// writable by its owner and by nobody else.
+func requireOwnerOnly(t *testing.T, path string) {
+	t.Helper()
+
+	info, err := os.Stat(path)
+	require.NoError(t, err, "%s must exist", path)
+	assert.Equal(
+		t,
+		ownerOnly,
+		info.Mode().Perm(),
+		"%s holds credentials and must not be readable by "+
+			"anyone but its owner",
+		path,
+	)
+}
+
+// requireDatabaseSetOwnerOnly asserts the mode of a database file and
+// of both WAL sidecars. The sidecars carry the same rows as the
+// database, so tightening only the main file fixes nothing.
+func requireDatabaseSetOwnerOnly(t *testing.T, dbPath string) {
+	t.Helper()
+
+	requireOwnerOnly(t, dbPath)
+	requireOwnerOnly(t, dbPath+"-wal")
+	requireOwnerOnly(t, dbPath+"-shm")
+}
+
+// TestMainDatabaseFilesAreOwnerOnly covers the tier the defect was
+// reported against: webhooker.db holds targets.config in plaintext —
+// bearer tokens, API keys, Slack webhook URLs — and the session
+// encryption key.
+func TestMainDatabaseFilesAreOwnerOnly(t *testing.T) {
+	t.Parallel()
+
+	lc := fxtest.NewLifecycle(t)
+
+	l, err := logger.New(lc, logger.LoggerParams{
+		Globals: &globals.Globals{
+			Appname: testAppname,
+			Version: testVersion,
+		},
+	})
+	require.NoError(t, err)
+
+	// A directory the application creates itself, not one t.TempDir
+	// made at 0700, so the mode below is the application's.
+	dataDir := filepath.Join(t.TempDir(), "data")
+
+	db, err := database.New(lc, database.DatabaseParams{
+		Config: &config.Config{DataDir: dataDir},
+		Logger: l,
+	})
+	require.NoError(t, err)
+
+	ctx := context.Background()
+	require.NoError(t, lc.Start(ctx))
+
+	defer func() { require.NoError(t, lc.Stop(ctx)) }()
+
+	// Write through the real model so the WAL is populated and both
+	// sidecars are on disk while the handle is open.
+	require.NoError(t, db.DB().Create(&database.Webhook{
+		Name: testWebhookName,
+	}).Error)
+
+	requireDatabaseSetOwnerOnly(
+		t, filepath.Join(dataDir, database.MainDBFileName),
+	)
+
+	// The data directory grants nothing to `other`. Asserted as a
+	// property rather than as an exact 0750, because MkdirAll applies
+	// the ambient umask: the exact mode is the developer's umask as
+	// much as the application's request, and pinning it would make
+	// `make check` pass or fail on where it is run. The group bits are
+	// deliberately left unasserted — deployments may rely on them.
+	info, err := os.Stat(dataDir)
+	require.NoError(t, err)
+	assert.Zero(
+		t,
+		info.Mode().Perm()&0o007,
+		"the data directory must not be world-accessible",
+	)
+}
+
+// TestPerWebhookEventDatabaseFilesAreOwnerOnly covers the events-*.db
+// tier. These carry no credential canaries since
+// https://git.eeqj.de/sneak/webhooker/issues/206, but they hold every
+// received request body and header.
+func TestPerWebhookEventDatabaseFilesAreOwnerOnly(t *testing.T) {
+	t.Parallel()
+
+	mgr, lc := setupTestWebhookDBManager(t)
+	ctx := context.Background()
+	require.NoError(t, lc.Start(ctx))
+
+	defer func() { require.NoError(t, lc.Stop(ctx)) }()
+
+	webhookID := uuid.New().String()
+
+	db, err := mgr.GetDB(webhookID)
+	require.NoError(t, err)
+
+	require.NoError(t, db.Create(&database.Event{
+		WebhookID:    webhookID,
+		EntrypointID: uuid.New().String(),
+		Method:       http.MethodPost,
+		Body:         "{}",
+	}).Error)
+
+	requireDatabaseSetOwnerOnly(t, mgr.DBPath(webhookID))
+}
+
+// TestArchiveDatabaseFilesAreOwnerOnly covers the archive-*.db tier.
+// internal/delivery builds that path and opens it through OpenSQLite,
+// the same single open path exercised here, so the mode is settled for
+// all three tiers in one place.
+func TestArchiveDatabaseFilesAreOwnerOnly(t *testing.T) {
+	t.Parallel()
+
+	ctx := context.Background()
+	path := filepath.Join(
+		t.TempDir(), "archive-"+uuid.New().String()+".db",
+	)
+
+	sqlDB, err := database.OpenSQLite(path, database.SQLiteModeCreate)
+	require.NoError(t, err)
+
+	defer func() { require.NoError(t, sqlDB.Close()) }()
+
+	_, err = sqlDB.ExecContext(ctx, "create table t (id integer)")
+	require.NoError(t, err)
+
+	requireDatabaseSetOwnerOnly(t, path)
+}
+
+// TestOpenSQLiteTightensFilesLeftWorldReadable is the upgrade case: a
+// data directory an earlier build left at 0644, including a
+// developer's own scratch directory, is fixed when it is opened rather
+// than staying exposed until it is recreated.
+func TestOpenSQLiteTightensFilesLeftWorldReadable(t *testing.T) {
+	t.Parallel()
+
+	dir := t.TempDir()
+	path := filepath.Join(dir, database.MainDBFileName)
+
+	// A database and both sidecars as the pre-fix build left them.
+	for _, p := range []string{path, path + "-wal", path + "-shm"} {
+		require.NoError(t, os.WriteFile(p, nil, 0o644)) //nolint:gosec // the mode under test
+	}
+
+	sqlDB, err := database.OpenSQLite(path, database.SQLiteModeCreate)
+	require.NoError(t, err)
+
+	require.NoError(t, sqlDB.Close())
+
+	requireDatabaseSetOwnerOnly(t, path)
+}
+
+// TestOpenSQLiteExistingModeDoesNotCreateTheFile guards the mechanism
+// the fix uses: OpenSQLite now creates the database file itself, and
+// must not do so for a caller that asked for an existing database. An
+// empty file materialized here would turn a missing-database error
+// into a silently empty one.
+func TestOpenSQLiteExistingModeDoesNotCreateTheFile(t *testing.T) {
+	t.Parallel()
+
+	ctx := context.Background()
+	path := filepath.Join(t.TempDir(), "absent.db")
+
+	sqlDB, err := database.OpenSQLite(path, database.SQLiteModeExisting)
+	if err == nil {
+		// sql.Open is lazy: force the connection that fails.
+		require.Error(t, sqlDB.PingContext(ctx))
+		require.NoError(t, sqlDB.Close())
+	}
+
+	_, statErr := os.Stat(path)
+	assert.ErrorIs(t, statErr, fs.ErrNotExist)
+}
+
+// TestReopenAfterRestartKeepsFilesOwnerOnly is the restart case: a
+// process that closed its files must be able to open them again at
+// 0600, including through a gorm handle, and the sidecars must come
+// back at 0600 too rather than at SQLite's own default.
+func TestReopenAfterRestartKeepsFilesOwnerOnly(t *testing.T) {
+	t.Parallel()
+
+	ctx := context.Background()
+	dir := t.TempDir()
+	path := filepath.Join(dir, database.MainDBFileName)
+
+	first, err := database.OpenSQLite(path, database.SQLiteModeCreate)
+	require.NoError(t, err)
+
+	_, err = first.ExecContext(ctx, "create table t (id integer)")
+	require.NoError(t, err)
+	require.NoError(t, first.Close())
+
+	second, err := database.OpenSQLite(path, database.SQLiteModeCreate)
+	require.NoError(t, err)
+
+	defer func() { require.NoError(t, second.Close()) }()
+
+	_, err = second.ExecContext(ctx, "insert into t (id) values (1)")
+	require.NoError(t, err)
+
+	requireDatabaseSetOwnerOnly(t, path)
+
+	var got int
+
+	require.NoError(t,
+		second.QueryRowContext(ctx, "select id from t").Scan(&got))
+	assert.Equal(t, 1, got)
+}

internal/database/sqlite_open.go

@@ -2,8 +2,11 @@ package database
 
 import (
 	"database/sql"
+	"errors"
 	"fmt"
+	"io/fs"
 	"net/url"
+	"os"
 	"time"
 
 	_ "modernc.org/sqlite" // Pure Go SQLite driver
@@ -72,6 +75,90 @@ const (
 	sqliteConnMaxIdleTime = time.Minute
 )
 
+// SQLiteFilePerm is the mode every SQLite file this service owns is
+// created with and held at: owner read/write, nothing for group or
+// other.
+//
+// These files hold credentials in plaintext. The main database stores
+// `targets.config` — bearer tokens, API keys, Slack webhook URLs — and
+// the session encryption key. SQLite left to itself creates them 0644
+// (see reserveSQLiteFile), which made the 0750 data directory the only
+// barrier; a bind-mounted directory supplied at 0755 removes it and
+// every local user on the host can read every stored credential.
+//
+// This is a file-mode fix and not encryption at rest. An unattended
+// process needs a key it can read without a human, so the key lands
+// beside the data and an attacker who can read the database can read
+// it too. See https://git.eeqj.de/sneak/webhooker/issues/212.
+const SQLiteFilePerm fs.FileMode = 0o600
+
+// reserveSQLiteFile puts path at SQLiteFilePerm before the driver ever
+// touches it, and tightens any sidecar already on disk.
+//
+// The mode has to be settled here rather than by a chmod after opening,
+// because SQLite picks it: robust_open substitutes
+// SQLITE_DEFAULT_FILE_PERMISSIONS (0644) whenever it is handed mode 0,
+// and findCreateFileMode yields 0 for a main database opened by URI
+// with no `modeof` parameter. A chmod afterwards would leave a window
+// in which the credentials are on disk world-readable.
+//
+// Creating the file ourselves also settles the sidecars, which is the
+// half that could quietly not work. SQLite does not create those at a
+// mode we choose — it derives both from the main database file:
+// `-wal` through findCreateFileMode, which stats the path with the
+// suffix stripped, and `-shm` in unixOpenSharedMemory from an fstat of
+// the already-open database descriptor. A main file at 0600 therefore
+// produces sidecars at 0600. A zero-length file is a valid empty
+// database, so reserving it changes nothing else.
+//
+// create says whether the caller is opening in a mode that may create
+// the database. When it is false a missing file is left missing, so
+// SQLite still reports the absence rather than this function
+// materializing an empty database the caller asked not to create.
+//
+// Chmod of a file that already exists is what tightens a data
+// directory an earlier build left at 0644 — including a developer's
+// own scratch directory — without any migration machinery.
+func reserveSQLiteFile(path string, create bool) error {
+	if create {
+		// gosec G304: the path is the database file the caller asked
+		// to open, and the driver is about to open the same path
+		// anyway. Creating it here is what fixes its mode.
+		f, err := os.OpenFile( //nolint:gosec // see above
+			path, os.O_RDWR|os.O_CREATE, SQLiteFilePerm,
+		)
+		if err != nil {
+			return fmt.Errorf("creating %s: %w", path, err)
+		}
+
+		err = f.Close()
+		if err != nil {
+			return fmt.Errorf("closing %s: %w", path, err)
+		}
+	}
+
+	// O_CREATE leaves an existing file's mode alone, and umask can only
+	// have narrowed a new one. Chmod settles both cases at exactly
+	// SQLiteFilePerm.
+	for _, p := range append(
+		[]string{path}, sqliteSidecarPaths(path)...,
+	) {
+		err := os.Chmod(p, SQLiteFilePerm)
+		if err != nil && !errors.Is(err, fs.ErrNotExist) {
+			return fmt.Errorf("securing %s: %w", p, err)
+		}
+	}
+
+	return nil
+}
+
+// sqliteSidecarPaths returns the files SQLite maintains beside a
+// database under WAL. They carry the same rows as the database itself,
+// so a fix that tightens only the main file has fixed nothing.
+func sqliteSidecarPaths(path string) []string {
+	return []string{path + "-wal", path + "-shm"}
+}
+
 // SQLiteDSN builds the connection string for one database file.
 //
 // mode is the SQLite URI open mode: "rwc" to create the file when it
@@ -138,9 +225,17 @@ func SQLiteDSN(path, mode string) string {
 // durability settings and pool bounds applied. mode is the SQLite URI
 // open mode ("rwc" or "rw").
 //
+// The file and its WAL sidecars are settled at SQLiteFilePerm before
+// the driver sees the path; see reserveSQLiteFile.
+//
 // The handle is returned rather than a *gorm.DB because the callers
 // wrap it in gorm themselves with their own logger.
 func OpenSQLite(path, mode string) (*sql.DB, error) {
+	err := reserveSQLiteFile(path, mode == SQLiteModeCreate)
+	if err != nil {
+		return nil, err
+	}
+
 	sqlDB, err := sql.Open("sqlite", SQLiteDSN(path, mode))
 	if err != nil {
 		return nil, fmt.Errorf(

internal/server/http.go

@@ -70,7 +70,7 @@ func (s *Server) serveUntilShutdown() {
 	err := s.httpServer.ListenAndServe()
 	if err != nil && !errors.Is(err, http.ErrServerClosed) {
 		s.log.Error("listen error", "error", err)
-		s.shutdownOnListenFailure()
+		s.shutdownWithFailure()
 	}
 }
 

internal/server/listen_failure_test.go

@@ -93,7 +93,7 @@ func requireListenFailureExit(t *testing.T, env *testEnv) {
 	select {
 	case sig := <-app.Wait():
 		require.Equal(
-			t, server.ListenFailureExitCode, sig.ExitCode,
+			t, server.StartupFailureExitCode, sig.ExitCode,
 			"listen failure must exit non-zero",
 		)
 	case <-time.After(listenFailureDeadline):

internal/server/sentry_failure_test.go

@@ -0,0 +1,99 @@
+package server_test
+
+import (
+	"context"
+	"net"
+	"strconv"
+	"testing"
+	"time"
+
+	"github.com/stretchr/testify/require"
+	"go.uber.org/fx"
+	"sneak.berlin/go/webhooker/internal/config"
+	"sneak.berlin/go/webhooker/internal/globals"
+	"sneak.berlin/go/webhooker/internal/server"
+)
+
+// TestSentryInitFailure_ShutsDownTheApp pins that error reporting
+// which is configured and cannot be started ends the application
+// instead of serving without it.
+//
+// The measured defect logged `sentry init failure` and kept running,
+// so the deployment served traffic with reporting off while every
+// other signal — SENTRY_DSN still set, the startup summary's own
+// field — said it was on. Nothing later in the process can notice
+// that reports are going nowhere, which is why this exits rather than
+// degrades.
+//
+// The DSN is placed on a hand-built Config, which is the only way to
+// reach this branch at all: loadFromEnv now parses SENTRY_DSN with
+// sentry.NewDsn, the same call sentry.Init makes, so a DSN that
+// survives configuration cannot fail initialisation in the SDK
+// version this pins. The branch stays because that is a property of
+// the SDK's current implementation rather than of its contract.
+func TestSentryInitFailure_ShutsDownTheApp(t *testing.T) {
+	t.Parallel()
+
+	port := freePort(t)
+
+	env := newTestEnvWithConfig(t, &config.Config{
+		DataDir:     t.TempDir(),
+		Environment: config.EnvironmentDev,
+		BindAddress: loopbackV4,
+		Port:        port,
+		SentryDSN:   "not-a-dsn",
+	})
+
+	app := fx.New(
+		fx.NopLogger,
+		fx.Supply(env.log, env.cfg, env.mw, env.hnd),
+		fx.Provide(globals.New, server.New),
+		fx.Invoke(func(*server.Server) {}),
+	)
+
+	startCtx, cancelStart := context.WithTimeout(
+		context.Background(), lifecycleTimeout,
+	)
+	defer cancelStart()
+
+	require.NoError(t, app.Start(startCtx))
+
+	select {
+	case sig := <-app.Wait():
+		require.Equal(
+			t, server.StartupFailureExitCode, sig.ExitCode,
+			"a sentry failure must exit non-zero",
+		)
+	case <-time.After(listenFailureDeadline):
+		t.Fatal("a sentry failure left the app running")
+	}
+
+	// The stop sequence still has to complete: the failure must reach
+	// shutdown through fx rather than around it.
+	stopCtx, cancelStop := context.WithTimeout(
+		context.Background(), lifecycleTimeout,
+	)
+	defer cancelStop()
+
+	require.NoError(t, app.Stop(stopCtx))
+
+	// And it must give up before it listens. A process that bound the
+	// port and then exited would have accepted requests it could not
+	// report on, which is the state under test in miniature.
+	requireBindable(t, port)
+}
+
+// requireBindable asserts that the port is free, which it is only if
+// the server under test never claimed it.
+func requireBindable(t *testing.T, port int) {
+	t.Helper()
+
+	var listenCfg net.ListenConfig
+
+	listener, err := listenCfg.Listen(
+		t.Context(), "tcp",
+		net.JoinHostPort(loopbackV4, strconv.Itoa(port)),
+	)
+	require.NoError(t, err, "the server bound a port it then gave up")
+	require.NoError(t, listener.Close())
+}

internal/server/server.go

@@ -51,12 +51,13 @@ const (
 	minSentryFlush = 250 * time.Millisecond
 )
 
-// ListenFailureExitCode is the status the process exits with when the
+// StartupFailureExitCode is the status the process exits with when
-// HTTP listener cannot be established, or dies for a reason other
+// the serving goroutine gives up: the HTTP listener cannot be
-// than a requested shutdown. It must stay non-zero: systemd
+// established or dies for a reason other than a requested shutdown, or
-// `Restart=on-failure` and Docker's restart policies key off it, and a
+// error reporting is configured and cannot be started. It must stay
-// zero exit would read as a deliberate stop.
+// non-zero: systemd `Restart=on-failure` and Docker's restart policies
-const ListenFailureExitCode = 1
+// key off it, and a zero exit would read as a deliberate stop.
+const StartupFailureExitCode = 1
 
 // SentryFlushBudget reports how long the Sentry flush may run when
 // remaining is the time left on the fx stop context after the HTTP
@@ -135,11 +136,25 @@ func New(lc fx.Lifecycle, params ServerParams) (*Server, error) {
 }
 
 // Run configures Sentry and starts serving HTTP requests.
+//
+// A Sentry failure ends the application instead of listening. It runs
+// before the listener rather than after it so that the process never
+// binds a port it is about to give up.
 func (s *Server) Run() {
 	s.configure()
 
 	// logging before sentry, because sentry logs
-	s.enableSentry()
+	err := s.enableSentry()
+	if err != nil {
+		s.log.Error(
+			"SENTRY_DSN is set but error reporting could not be "+
+				"started; refusing to serve with it off",
+			"error", err,
+		)
+		s.shutdownWithFailure()
+
+		return
+	}
 
 	s.serve()
 }
@@ -150,11 +165,23 @@ func (s *Server) MaintenanceMode() bool {
 	return s.params.Config.MaintenanceMode
 }
 
-func (s *Server) enableSentry() {
+// enableSentry initialises the Sentry SDK when error reporting is
+// configured, and reports the failure when it is configured and cannot
+// be initialised. A DSN that is not set is not a failure: reporting
+// stays off and the server starts normally.
+//
+// There is no fallback to running with reporting off. An operator who
+// set SENTRY_DSN asked for failures to be visible, and serving traffic
+// with reporting quietly off is the one state nothing can ever tell
+// them about — the DSN is still set, so every later signal says it is
+// on. Config already refused a DSN the SDK cannot parse, which is what
+// a typo produces, so reaching this branch means the SDK refused
+// something that parsed: not a condition to guess at either.
+func (s *Server) enableSentry() error {
 	s.sentryEnabled.Store(false)
 
-	if s.params.Config.SentryDSN == "" {
+	if !s.params.Config.SentryEnabled() {
-		return
+		return nil
 	}
 
 	err := sentry.Init(sentryClientOptions(
@@ -166,19 +193,19 @@ func (s *Server) enableSentry() {
 		),
 	))
 	if err != nil {
-		s.log.Error("sentry init failure", "error", err)
+		return fmt.Errorf("initialising sentry: %w", err)
-		// Don't use fatal since we still want the service to run
-		return
 	}
 
 	s.log.Info("sentry error reporting activated")
 	s.sentryEnabled.Store(true)
+
+	return nil
 }
 
 // serve installs the signal watcher, starts the listener and blocks
 // until the server's context is cancelled. The process exit status is
 // fx's to decide — from a signal, or from the code
-// shutdownOnListenFailure hands the Shutdowner — so this reports
+// shutdownWithFailure hands the Shutdowner — so this reports
 // nothing back to its caller.
 func (s *Server) serve() {
 	ctx, cancelFunc := context.WithCancel(context.Background())
@@ -208,20 +235,24 @@ func (s *Server) serve() {
 	// Do not call cleanShutdown() here to avoid double invocation.
 }
 
-// shutdownOnListenFailure ends the application after the HTTP
+// shutdownWithFailure ends the application non-zero from the serving
-// listener failed. The fx OnStart hook returns as soon as the serving
+// goroutine. It is how anything on that goroutine fails fatally: the
-// goroutine is spawned, so nothing downstream of it ever learns that
+// fx OnStart hook returns as soon as the goroutine is spawned, so
-// the listen failed: fx reports RUNNING and the process sits alive
+// nothing downstream of it ever learns that the goroutine gave up. fx
-// with nothing bound, which is invisible to systemd and Docker
+// reports RUNNING and the process sits alive having done neither what
-// restart policies. Asking the Shutdowner to stop the app with a
+// it was asked nor anything visible instead, which systemd and
-// non-zero code is what turns that into a visible failure.
+// Docker restart policies cannot see. Asking the Shutdowner to stop
+// the app with a non-zero code is what turns that into a visible
+// failure, and it is the whole of "fatal" here — no panic, no
+// os.Exit, and every stop hook still runs.
 //
-// The context cancel that follows only unwinds serve()'s own wait.
+// The context cancel that follows only unwinds serve()'s own wait,
-// The shutdown itself runs through fx's normal stop sequence, so the
+// and is skipped before serve has installed one. The shutdown itself
-// clean-shutdown drain in cleanShutdown is reached unchanged.
+// runs through fx's normal stop sequence, so the clean-shutdown drain
-func (s *Server) shutdownOnListenFailure() {
+// in cleanShutdown is reached unchanged.
+func (s *Server) shutdownWithFailure() {
 	err := s.params.Shutdowner.Shutdown(
-		fx.ExitCode(ListenFailureExitCode),
+		fx.ExitCode(StartupFailureExitCode),
 	)
 	if err != nil {
 		s.log.Error("shutdown request failed", "error", err)