Security policy · LMS-13
Logging and Monitoring Standard
1. Purpose and scope
This standard defines how Enigmatech Mexico SAPI de CV ("Enigmatech") records, protects, reviews and alerts on activity in the EnigmaSuite Platform and the systems that support it, so that misuse of data, including data received from Amazon and other partner platforms, is detected and investigated quickly. It supports the Information Security Policy.
2. What is logged
Across every channel (application APIs, database and storage APIs, administrative consoles and servers), logs record:
- authentication: sign-ins, failed sign-ins, multi-factor challenges, password changes and session revocations;
- authorization: role and permission changes, access denied, and use of privileged accounts;
- data access and changes: creation, modification, export and deletion of records, including access to buyer personal information and calls to partner platform APIs;
- configuration and production changes: deployments, changes to firewall, DNS, secrets and integrations;
- system errors and security alerts.
Each entry records the date and time, the account or service, the action, the target, the source and whether it succeeded or failed.
3. Protecting logs
- No personal information in logs. Logs identify records by their internal identifiers, not by buyer names, addresses, phone numbers or emails. Secrets and tokens are never logged. A log that is found to contain personal information is purged and the cause is fixed.
- Tamper resistance. Audit records are append-only for application users, are written with server-side timestamps, and can be read only by authorized administrators. Infrastructure logs are kept by the providers outside the reach of application code.
- Centralization and time. Security events are gathered in our security operations module (Enigma Shield). All systems use synchronized time (UTC).
- Retention. Security and audit logs are kept for at least 12 months, and up to 24 months, or longer while an investigation is open.
4. Monitoring and alarms
Automated detection rules raise alerts in Enigma Shield, which deduplicates them and opens an incident with a response deadline. Alarms cover at least:
- repeated unauthorized or failed calls, and accounts locked by failed sign-ins;
- request rates and data retrieval volumes outside the expected range for a customer, a user or an integration, including unusual volumes of buyer personal information read from Amazon;
- access to canary records: decoy records that no legitimate process reads;
- any access with privileged accounts, and access outside business hours or from unexpected locations;
- every production change and every change to an integration's credentials;
- signs that our data or credentials appear outside our systems (for example, secret-scanning alerts or public leak reports).
5. Review
- Alerts are triaged as they arrive and at least once every business day.
- The Security Lead reviews the security logs and alert trends at least every two weeks, and records the review.
- Every alarm investigation is documented and, if it reveals an incident, handled under the Incident Response Policy.
6. Health monitoring
Health checks watch the availability of the application, database, queues and scheduled jobs. A job that fails repeatedly alerts the engineering team; a job that stops reporting is treated as a failure, not as success.
7. Review of this standard
This standard is reviewed at least once a year and after any incident in which logging or monitoring did not work as expected.