PROPERTIES.md

Duty Properties

Duty uses runtime properties for behavior that should be adjustable without changing code. These properties are plain string key/value pairs. They are separate from model properties fields such as config_items.properties, components.properties, or connection properties, which are resource metadata stored in JSON columns.

The machine-readable schema for JSON/YAML maps of these runtime properties is PROPERTIES.schema.json. The schema is intentionally kept outside schema/openapi because that directory is generated.

Setting Properties

CLI

Applications that bind github.com/flanksource/commons/properties.BindFlags accept -P / --properties:

duty-command -P log.level=debug -P query.log=true
duty-command --properties log.level.http=trace

CLI properties are held in the process-local commons property store.

Environment

Duty does not provide a generic environment-variable-to-property mapper for runtime properties. To set a runtime property from the environment, pass it through the CLI, a properties file, DB, or code that calls properties.Set.

duty-command -P query.log="$QUERY_LOG"

Startup configuration has explicit environment variables and environment indirection:

Env var	Purpose
DB_URL	Default value for --db if --db is not set.
PGRST_JWT_SECRET	Default value for --postgrest-jwt-secret if the flag is not set.
PGRST_VERSION	Overrides the bundled PostgREST version.
PGRST_ARCH	Overrides the PostgREST binary architecture.
OTEL_EXPORTER_OTLP_ENDPOINT	Default OpenTelemetry collector endpoint.
OTEL_LABELS	Comma-separated OpenTelemetry resource labels, key=value,key2=value2.
POD_NAMESPACE	Kubernetes namespace used by leader election.
MC_HOSTNAME_OVERRIDE	Hostname override used by leader election.
DUTY_DB_DISABLE_RLS	Used by tests and hack/migrate; disables RLS when set to true.
DUTY_DB_URL, DUTY_DB_CREATE, DUTY_DB_DATA_DIR	Test database setup only.
DUTY_BENCH_SIZES	Benchmark size list only.
KUBECONFIG	Kubernetes config path used by the Kubernetes client.

String startup flags are also passed through api.Config.ReadEnv(): when a flag value is the name of an environment variable, Duty uses that variable's value. For example, --db DUTY_DB_URL reads DUTY_DB_URL.

File

Code may call properties.LoadFile("duty.properties"). The file format is:

# comments are allowed
log.level=debug
query.log=true
topology.query.timeout=45s

The commons property loader watches the loaded file and reloads it on changes. There is no built-in --properties-file flag in this package.

Known embedding applications load these default files:

Application	Default file
incident-commander / mission-control	mission-control.properties
config-db	config-db.properties
canary-checker	canary-checker.properties

Database

Context-aware properties are read from the properties table:

INSERT INTO properties (name, value)
VALUES ('query.log', 'true')
ON CONFLICT (name) DO UPDATE SET value = excluded.value;

From Go, use:

context.UpdateProperty(ctx, "query.log", "true")
context.UpdateProperties(ctx, map[string]string{"query.log": "true"})

Database properties are cached in process for 15 minutes. UpdateProperty and UpdateProperties clear the cache after writing.

Annotations

For object-scoped lookup through ctx.Properties(), add annotations with one of these prefixes:

metadata:
  annotations:
    mission-control/query.log: "true"
    canary-checker/topology.query.timeout: 45s

The prefix is stripped before lookup, so mission-control/query.log sets query.log. Child objects override parent objects.

Logging annotations also accept unprefixed forms in addition to the prefixed forms:

metadata:
  annotations:
    log.level: debug
    trace: "true"
    debug: "true"

Precedence

For context-aware properties resolved by ctx.Properties():

1. Process-local commons properties: CLI -P, loaded properties file, or code that calls properties.Set.
2. Object annotations, with child object annotations overriding parent object annotations.
3. Database rows in the properties table.
4. The hard-coded default at the call site.

Process-global properties that call properties.String, properties.Int, properties.Duration, or properties.On directly do not read DB rows or annotations. They only see the process-local commons property store.

Boolean values for ctx.Properties().On are true when set to true, enabled, or on. ctx.Properties().Off treats false, disabled, and off as off. The lower-level commons properties.On only treats true as true.

Startup Flags

These are not runtime properties, but they are the main Duty startup settings:

Flag	Default	Notes
--db	DB_URL	PostgreSQL connection string. The default is resolved from DB_URL.
--db-schema	public	PostgreSQL schema.
--postgrest-uri	http://localhost:3000	Localhost starts an embedded PostgREST process. Empty disables PostgREST.
--postgrest-log-level	info	PostgREST log level.
--postgrest-jwt-secret	PGRST_JWT_SECRET	JWT secret. The default is resolved from PGRST_JWT_SECRET.
--disable-postgrest	varies	Deprecated; use --postgrest-uri ''.
--postgrest-role	postgrest_api	Authenticated PostgREST database role.
--postgrest-anon-role	postgrest_anon	Unauthenticated PostgREST database role.
--postgrest-max-rows	2000	Hard row limit for PostgREST.
--db-log-level	error	GORM log level: trace, debug, info, error.
--disable-kubernetes	false	Disable Kubernetes integration.
--db-metrics	false	Register GORM Prometheus metrics.
--skip-migrations	mode-dependent	Skip database migrations when migrations run by default.
--db-migrations	mode-dependent	Run migrations when migrations are skipped by default. Deprecated in run-by-default mode.
--otel-collector-url	OTEL_EXPORTER_OTLP_ENDPOINT	OpenTelemetry gRPC collector endpoint.
--otel-service-name	caller supplied	OpenTelemetry service name.
--otel-insecure	true	Disable TLS for the OpenTelemetry collector.

Context-Aware Properties

These properties are read through ctx.Properties() and can be set via CLI, file, DB, or annotations.

Property	Type	Default	Effect
artifacts.connection	string	empty	Connection URL for external artifact/blob storage. Empty uses inline DB-backed blob storage.
casbin.auto.save	bool	true	Enables Casbin auto-save.
casbin.cache	bool	true	Enables the Casbin enforcer cache.
casbin.cache.expiry	duration	1m	Casbin cache expiry.
casbin.cache.reload.interval	duration	5m	Casbin policy auto-load interval.
casbin.explain	bool	false	Uses Casbin EnforceEx and logs matched rules.
casbin.log.level	int	1	Enables Casbin logging when >= 2.
db.connection.timeout	duration	1h	Statement timeout applied to the application DB user role.
db.postgrest.timeout	duration	1m	Statement timeout applied to PostgREST DB roles.
envvar.cache.timeout	duration	5m	Cache TTL for Kubernetes Secret and ConfigMap env var lookups.
envvar.helm.cache.timeout	duration	envvar.cache.timeout	Cache TTL for Helm release value lookups.
envvar.lookup.timeout	duration	5s	Timeout for resolving an EnvVar from Kubernetes sources.
har.captureContentTypes	CSV string	empty	Restricts HAR content capture to matching content types.
har.maxBodySize	int bytes	65536	Maximum HAR body capture size.
job.ResetIsPushed.ignore_deleted_at	bool	false	When true, reset-is-pushed queries do not add deleted_at IS NULL.
job.ResetIsPushed.interval_days	int	7	Lookback window for resetting is_pushed.
job.eviction.period	duration	1m	Sleep period for job-history eviction when no eviction IDs are queued.
job.jitter.disable	bool	false	Disables schedule jitter for periodic jobs.
leader.lease.duration	duration	30s	Kubernetes leader election lease duration.
log.level	string	logger default	Raises effective context observability level globally.
log.level.http	string	unset	Enables HTTP request/response header logging at debug; includes bodies at trace.
log.level.http.har	string	unset	Enables HAR capture for HTTP at debug; includes full body capture at trace.
log.level.<feature>	string	unset	Raises effective context logging for a named feature. For Kubernetes, aliases are kubernetes, kubectl, and k8s.
log.level.<feature>.har	string	unset	Raises effective HAR capture level for a named feature.
log.level.resourceSelector	string	unset	When set, logs generated resource selector SQL at trace level on the resourceSelector logger.
postgres.session.<setting>	string	unset	Applied by ApplySessionProperties as SET LOCAL <setting> = '<value>' inside a transaction.
query.log	bool	false	Logs resource selector and query logger output at normal verbosity.
secretkeeper.cache.ttl	duration	10m	TTL for the cloud secret keeper cache.
shell.connection.wait_before_cleanup	duration	0	Wait before cleaning up shell connection artifacts.
topology.cache.age	duration	5m	Cache age for topology responses.
topology.query.timeout	duration	30s	Default topology query timeout when the context has no deadline.
update_is_pushed.batch.size	int	200	Batch size for marking pushed records during upstream reconciliation.
upstream.client.cache.view-columns.duration	duration	go-cache default	Cache duration for upstream view-column client lookups.
view.http.body.max_size_bytes	int bytes	26214400	Maximum HTTP response body size for HTTP data queries. Non-positive values fall back to the default.

Mission Properties

These are additional properties used by ../incident-commander. Unless noted as process-global in the later table, they are resolved through ctx.Properties() and can come from CLI/file, DB, or annotations.

Property	Type	Default	Effect
access.log	bool	true	Enable access logging.
access.log.colors	bool	true	Enable colors in detailed access logs.
access.log.debug	bool	false	Enable debug logging for access log middleware.
access.log.request.body	bool	false	Include request bodies in access logs.
access.log.request.body.max	int bytes	2048	Maximum request body bytes captured by access logs.
access.log.request.header	bool	mixed defaults	Include request headers in access logs.
access.log.request.id	bool	false	Include request IDs in access logs.
access.log.response.body	bool	false	Include response bodies in access logs.
access.log.response.body.max	int bytes	8192	Maximum response body bytes captured by access logs.
access.log.skip.sanitize	bool	false	Skip access log sanitization.
access.log.spanId	bool	true	Include span IDs in access logs.
access.log.traceId	bool	true	Include trace IDs in access logs.
access.log.userAgent	bool	false	Include user-agent values in access logs.
artifacts.max_read_size	int bytes	52428800	Maximum artifact bytes read for playbook artifact responses. Values <= 0 disable the guard.
auth.impersonation	bool/off switch	false	off, false, or disabled disables scope impersonation.
dashboard.default.view	string	mission-control-dashboard	Default dashboard view name or namespace/name.
event_queue.maxAge	duration	720h	Maximum age for event_queue rows before cleanup.
events.audit.size	int	20	Number of recent events retained in audit rings.
<event>.batchSize	int	handler value	Batch size for a named async event consumer, for example notification.send.batchSize.
<event>.debug	bool/off switch	false	Enables debug logging for a named async event consumer when set to off/false by current code.
<event>.trace	bool/off switch	false	Enables trace logging for a named async event consumer when set to off/false by current code.
incidents.disable	bool	false	Disable incident notification behavior.
job.history.agentItemsToRetain	int	3	Agent job-history entries to retain per status grouping.
job.history.maxAge	duration	720h	Maximum job history age before cleanup.
job.history.running.maxAge	duration	4h	Maximum running job age before marking stale.
mcp.template.max-length	int bytes	65536	Maximum MCP template size.
mcp.template.timeout	duration	10s	MCP template rendering timeout.
metrics.agents.cache_ttl	duration	5m	Prometheus agent collector cache TTL.
metrics.canaries.cache_ttl	duration	5m	Prometheus canary collector cache TTL.
metrics.checks.cache_ttl	duration	5m	Prometheus check collector cache TTL.
metrics.checks.labels	CSV string	empty	Check label include/exclude patterns for metrics.
metrics.config_items.cache_ttl	duration	5m	Prometheus config item collector cache TTL.
metrics.disable	CSV string	empty	Metric names to disable. * disables all supported metrics.
metrics.prefix	string	empty	Metric name prefix.
notification.max-retries	int	4	Maximum notification delivery retries.
notifications.dedup.window	duration	24h	Notification de-duplication window.
notifications.error_reset_duration	duration	1h	How long before notification errors can be reset.
notifications.group_by_interval	duration	24h	Default interval for grouped notifications.
notifications.max.count	int	50	Maximum notifications per rate-limit window.
notifications.max.window	duration	4h	Notification rate-limit window.
playbook.action.ai.log-prompt	bool	false	Log AI action prompts.
playbook.action.consumers	int	5	Number of playbook action consumers.
playbook.consumer.timeout	duration	1m	Playbook consumer timeout.
playbook.retention.age	duration	720h	Retention period for deleted playbooks.
playbook.run.timeout	duration	30m	Default playbook run timeout.
playbook.runner.disabled	bool	false	Disable playbook action runners.
playbook.runner.longpoll.timeout	duration	45s	Long-poll timeout for remote playbook runners.
playbook.scheduler.disabled	bool	false	Disable playbook run scheduler.
playbook.schedulers	int	5	Number of playbook run schedulers.
rls.debug	bool	false	Log RLS payloads.
rls.disable	bool	false	Disable RLS in startup checks.
rls.enable	bool	false	Enable RLS.
scope.cache.ttl	duration	1m	RBAC scope cache TTL.
settings.user.disabled	bool	false	Set by auth middleware when the current user is disabled.
shorturl.defaultExpiry	duration	2160h	Default short URL expiry.
slack.max-url-length	int	50	Maximum Slack URL length before shortening. Values above 3000 are ignored by code.
upstream.pull_playbook_actions	bool	true	Schedule upstream playbook action pull jobs.
view.refresh.max-timeout	duration	1m	Maximum timeout for asynchronous view refreshes.

Config DB Properties

These are additional properties used by ../config-db.

Property	Type	Default	Effect
azuredevops.concurrency	int	5	Azure DevOps scraper concurrency.
azuredevops.pipeline.max_age	duration	168h	Maximum Azure DevOps pipeline run age to scrape.
azuredevops.terminal_cache.ttl	duration	1h	Azure DevOps terminal status cache TTL.
change_retention.delete_batch_size	int	1000	Batch size for config change retention deletes.
changes.dedup.disable	bool	false	Disable config change de-duplication.
changes.dedup.window	duration	1h	Config change de-duplication window.
config.retention.period	duration	168h	Retention period for deleted config items.
config.retention.stale_item_age	duration	scraper default	Age after which stale config items are soft deleted.
config_analysis.retention.max_age	duration	48h	Age after which stale config analyses are marked resolved.
config_analysis.set_status_closed_days	int days	7	Days after which resolved config analyses are closed.
config_scraper.retention.period	duration	720h	Retention period for deleted config scraper records.
diff.rust-gen	bool	false	Use the alternate diff implementation when available.
external.cache.timeout	duration	24h	External entity cache timeout.
incremental_scrape_event.lag_threshold	duration	30s	Slow-event threshold for incremental scrape event logging.
kubernetes.get.concurrency	int	10	Concurrency for Kubernetes fetch operations.
kubernetes.rbac_config_access	bool	true	Enable config access generation from Kubernetes RBAC.
scraper.concurrency	int	12	Global config scraper concurrency.
scraper.<type>.concurrency	int	type default	Per-type scraper concurrency. Known types include aws, azure, azuredevops, file, gcp, githubactions, http, kubernetes, kubernetesfile, slack, sql, terraform, trivy, and playwright.
scraper.<uid>.schedule	string	spec/default	Per-scrape-config schedule override by UID.
scraper.<type>.schedule.min	duration	29s	Minimum schedule interval for a scraper type.
scraper.aws.trusted_advisor.minInterval	duration	16h	Minimum interval between AWS Trusted Advisor scrapes.
scraper.diff.disable	bool	false	Disable config diff generation.
scraper.diff.timer.minSize	int bytes	20480	Minimum config size before diff memory timing at high verbosity.
scraper.log.items	bool	false	Log config scraper item processing details.
scraper.log.slow_diff_threshold	duration	1s	Threshold for slow diff warnings.
scraper.timeout	duration	4h	Default scraper timeout.
scraper.<key>	bool	varies	ScrapeContext.PropertyOn prefixes keys with scraper. and also checks scraper.<uid>.<key>. Common keys include azure.devops.incremental, capture.har, capture.logs, capture.snapshots, runNow, disable, watch.disable, log.exclusions, log.skipped, log.noResourceId, log.items, log.missing, log.relationships, log.rule.expr, log.transforms, and log.changes.unmatched.
scraper.scraper.label.missing	bool	false	Effective key for current PropertyOn("scraper.label.missing") usage.
scraper.scraper.tag.missing	bool	false	Effective key for current PropertyOn("scraper.tag.missing") usage.
scrapers.default.schedule	string	startup flag default	Default schedule for scrape configs without an explicit schedule.
scrapers.event.stale-timeout	duration	1h	Scraper event stale timeout.
scrapers.event.workers	int	2	Number of scraper event workers.
scrapers.githubactions.concurrency	int	10	GitHub Actions API request concurrency per repository.
scrapers.githubactions.maxAge	duration	168h	Maximum age for GitHub Actions workflow runs.

Canary Checker Properties

These are additional properties used by ../canary-checker.

Property	Type	Default	Effect
canary.retention.age	duration	168h	Retention period for soft-deleted canaries.
canary.status.max.error	int bytes	131072	Maximum check status error length.
canary.status.max.message	int bytes	4096	Maximum check status message length.
check.*.disabled	bool	false	Disable canary check job synchronization.
check.retention.age	duration	168h	Retention period for soft-deleted checks.
check.status.retention.days	int days	30	Check status retention in days.
checks.kubernetesResource.maxResources	int	10	Maximum total Kubernetes resources allowed in a Kubernetes resource check.
component.retention.period	duration	168h	Retention period for soft-deleted components.
components.delete_batch_size	int	100	Batch size for component deletion during topology sync.
http.har	bool	false	Enable HAR collection.
http.har.location	string	.	Directory where HAR files are written.
pubsub.max_messages	int	1000	Maximum Pub/Sub messages read by a canary check.
s3.list.max-objects	int	50000	Maximum S3 objects listed by folder checks.
upstream.pull_canaries	bool	true	Schedule canary upstream pull jobs.

Flanksource UI Property Usage

../flanksource-ui reads /properties as feature flags. These are database properties served by the API; they are not object annotations. The Settings > Feature Flags page can create and update DB-backed rows, but rows with source=local are displayed read-only.

The UI also uses many resource metadata properties fields, for example connection form properties, topology/config display properties, and playbook parameter UI hints such as language, jsonSchemaUrl, options, filter, multiline, min, max, minLength, maxLength, and regex. Those are not runtime properties and are not listed in the schema.

UI Feature Flags

Feature flags use the property name <feature>.disable. A feature is disabled only when the property value is exactly the string true; missing rows and any other value leave the feature enabled.

Property	Effect
topology.disable	Hide or disable topology UI surfaces.
health.disable	Hide or disable health UI surfaces.
incidents.disable	Hide or disable incident UI surfaces. Also disables incident notification behavior in incident-commander when read by the backend.
config.disable	Hide or disable config UI surfaces.
logs.disable	Hide or disable log UI surfaces.
playbooks.disable	Hide or disable playbook UI surfaces.
applications.disable	Hide or disable application UI surfaces.
views.disable	Hide or disable custom view UI surfaces.
ai.disable	Hide or disable AI actions and prompts in UI surfaces that check this flag.
agents.disable	Hide or disable agent UI surfaces.
settings.connections.disable	Hide or disable connection settings.
settings.users.disable	Hide or disable user settings.
settings.teams.disable	Hide or disable team settings.
settings.rules.disable	Hide or disable rules settings.
settings.config_scraper.disable	Hide or disable config scraper settings. Also disabled when config.disable=true.
settings.topology.disable	Hide or disable topology settings. Also disabled when topology.disable=true.
settings.health.disable	Hide or disable health settings. Also disabled when health.disable=true.
settings.job_history.disable	Hide or disable job history settings. Also disabled when health.disable=true.
settings.feature_flags.disable	Hide or disable the feature flags settings page.
settings.logging_backends.disable	Hide or disable logging backend settings.
settings.event_queue_status.disable	Hide or disable event queue status settings.
settings.organization_profile.disable	Hide or disable organization profile settings.
settings.notifications.disable	Hide or disable notification settings.
settings.playbooks.disable	Hide or disable playbook settings.
settings.integrations.disable	Hide or disable integration settings.
settings.permissions.disable	Hide or disable permission settings.
settings.artifacts.disable	Hide or disable artifact settings.

UI Snippets

Property	Type	Source	Effect
flanksource.ui.snippets	JavaScript function expression string	local only	Executed once after the authenticated user is available. The function receives { user, organization }.

Only a flanksource.ui.snippets row whose source is local is executed by the UI. DB-backed rows with the same name are fetched and visible in the feature-flag list, but the snippet hook ignores them.

Example value:

({ user, organization }) => {
  window.analytics?.identify(user?.id, {
    email: user?.email,
    organization: organization?.name
  });
}

UI Dashboard Selection

Property	Type	Effect
defaults.dashboard_view	string	Used by the UI sidebar to decide which custom view should be shown as the dashboard navigation item. It accepts a view UUID, namespace/name, or name.
dashboard.default.view	string	Used by the backend /api/dashboard endpoint to resolve the actual homepage dashboard view. It accepts namespace/name or name and defaults to mission-control-dashboard.

UI Proxy Selection

Property	Type	Effect
proxy.disable	bool	In Clerk auth mode, overrides the organization's direct metadata. When true, the UI bypasses the proxy and points API clients at the organization's backend_url.

Job Properties

Jobs support global, per-job, and per-job-id property names:

Property pattern	Type	Default	Effect
jobs.<name>.schedule	string	job value	Overrides a job's cron schedule.
jobs.<name>.<id>.schedule	string	job value	Intended per-id schedule override. The current lookup checks jobs.<name>.schedule first.
jobs.<name>.timeout	duration	job value	Overrides job timeout.
jobs.<name>.<id>.timeout	duration	job value	Intended per-id timeout override. The current lookup checks jobs.<name>.timeout first.
jobs.<name>.history	bool	true	Enables job history.
jobs.<name>.<id>.history	bool	true	Enables job history for a specific job id.
jobs.history	bool	true	Fallback for all jobs.
jobs.<name>.trace	bool	false	Enables trace logging for a job.
jobs.<name>.<id>.trace	bool	false	Enables trace logging for a specific job id.
jobs.trace	bool	false	Fallback for all jobs.
jobs.<name>.debug	bool	false	Enables debug logging for a job.
jobs.<name>.<id>.debug	bool	false	Enables debug logging for a specific job id.
jobs.debug	bool	false	Fallback for all jobs.
jobs.<name>.singleton	bool	job value	Overrides singleton behavior.
jobs.<name>.<id>.singleton	bool	job value	Overrides singleton behavior for a specific job id.
jobs.singleton	bool	job value	Fallback for all jobs.
jobs.<name>.disable	bool	false	Disables a job.
jobs.<name>.<id>.disable	bool	false	Disables a specific job id.
jobs.disable	bool	false	Fallback for all jobs.
jobs.<name>.disabled	bool	false	Alias for disable.
jobs.<name>.<id>.disabled	bool	false	Alias for disable for a specific job id.
jobs.disabled	bool	false	Fallback alias for all jobs.
jobs.<name>.retention.success	int	job value	Successful job-history entries to retain.
jobs.<name>.<id>.retention.success	int	job value	Intended per-id success retention override. The current lookup checks jobs.<name>.retention.success first.
jobs.<name>.retention.failed	int	job value	Failed, warning, or skipped job-history entries to retain.
jobs.<name>.<id>.retention.failed	int	job value	Intended per-id failed retention override. The current lookup checks jobs.<name>.retention.failed first.

Boolean job properties are looked up from most specific to least specific: jobs.<name>.<id>.<key>, jobs.<name>.<key>, then jobs.<key>.

String and int job helpers currently check jobs.<name>.<key> before jobs.<name>.<id>.<key>.

Process-Global Properties

These properties are read directly from the commons process-local property store. They can be set via CLI -P, a loaded properties file, debug property POSTs, or code that calls properties.Set, but not via DB rows or annotations.

Property	Type	Default	Effect
access_token.default_expiry	duration	2160h	Default expiry for generated access tokens.
canary.status.max.error	int bytes	131072	Maximum check status error length.
canary.status.max.message	int bytes	4096	Maximum check status message length.
change_retention.delete_batch_size	int	1000	Batch size for config change retention deletes.
components.delete_batch_size	int	100	Batch size for component deletion during topology sync.
config.traversal.cache_expiry.min	duration	2h	Minimum randomized expiry for config traversal cache entries.
config.traversal.cache_expiry.max	duration	4h	Maximum randomized expiry for config traversal cache entries.
config_analysis.set_status_closed_days	int days	7	Days after which resolved config analyses are closed.
db.migrate.skip	bool	false	Skips database migrations in migrate.Migrate.
diff.rust-gen	bool	false	Use the alternate diff implementation when available.
envvar.lookup.log	bool	false	Logs resolved env var lookup values when the logger verbosity is high enough.
external.cache.timeout	duration	24h	External entity cache timeout in config-db.
http.body.disabled	bool	false	Disables HTTP body capture in commons HTTP trace middleware.
http.headers.disabled	bool	false	Disables HTTP header capture in commons HTTP trace middleware.
http.log.response.body.length	int bytes	4096	Maximum logged HTTP response body length.
incremental_scrape_event.lag_threshold	duration	30s	Slow-event threshold for incremental scrape event logging.
job_history.agent_cleanup.batch_size	int	2000	Batch size for stale agent job-history cleanup.
kubernetes.cache.timeout	duration	240m	Kubernetes discovery cache timeout for the REST mapper.
log.color	bool	logger flag default	Enables colored logs. Automatically forced to false when log.json=true.
log.color.<logger>	bool	log.color	Color setting for a named logger.
log.caller	bool	logger flag default	Adds source caller information to logs.
log.caller.<logger>	bool	log.caller	Caller setting for a named logger.
db.log.level	string	unset	Updates the db logger level through the commons logger property listener.
log.db.maxLength	int	1024	Maximum SQL log length.
log.db.params	bool	false	Logs SQL parameters when DB trace logging is enabled.
log.db.slowThreshold	duration	1s	GORM slow query threshold.
log.json	bool	logger flag default	Emits JSON logs.
log.kubeproxy	bool	false	Enable kube proxy logging in incident-commander.
log.level	string	info	Root logger level for commons loggers.
log.level.<logger>	string	log.level	Named logger level. Special case: log.level.http wraps the default HTTP client transport with a logger.
log.report.caller	bool	logger flag default	Alias used by logger.Configure; also updates caller reporting.
log.time.format	string	15:04:05.000	Log timestamp format.
log.time.format.<logger>	string	log.time.format	Timestamp format for a named logger.
memory.stats	duration	0	When positive, starts periodic memory stats logging on the debug server.
metrics.auth.disabled	bool	false	Disable authentication for /metrics in incident-commander.
notification.tracing	bool	false	Enable notification tracing.
notifications.labels.order	string	built-in label order	Overrides default display ordering for selected labels.
notifications.labels.whitelist	string	built-in whitelist	Overrides default label whitelist groups.
pubsub.max_messages	int	1000	Maximum Pub/Sub messages read by a canary check.
response.strip_upstream_cors	bool	true	Strip upstream CORS headers in incident-commander proxy responses.
shell.allowed.envs	CSV string	empty	Additional environment variable names passed through to shell executions. Read during package init.
shell.jq.timeout	duration	5s	Timeout for jq execution.
shell.yq.timeout	duration	shell.jq.timeout	Timeout for yq execution.
smtp.debug	bool	false	Enable SMTP debug logging.
upstream.pull_canaries	bool	true	Schedule canary upstream pull jobs.
upstream.summary.fkerror_id_count	int	10	Number of foreign-key error IDs included in upstream reconciliation summaries.

Observability Notes

HTTP and HAR context properties are feature-aware. For example:

log.level.http=debug
log.level.http.har=trace
log.level.kubernetes=debug
log.level.kubernetes.har=debug
har.maxBodySize=131072

For plain HTTP logging, debug logs headers and trace logs headers plus bodies. For HAR capture, debug captures request/response metadata and trace enables the HAR collector middleware with body capture subject to HAR configuration.

Discovery Endpoints

When the debug routes are registered:

Endpoint	Result
GET /debug/properties	Supported context-aware properties that have been touched in the running process, including type, default, and current value.
GET /debug/system/properties	Process-local commons properties.
POST /debug/property	Sets a process-local commons property for the running process.
echo.Properties handler	Combined process-local and DB properties, with their source, wherever the embedding application mounts it.

GET /debug/properties is populated lazily by calls to ctx.Properties(). It is useful for introspection, but it is not a complete static registry until the relevant code paths have executed.