Skip to content

chore(observability): align the AI service log format with the web/worker logs #51

Description

@Fluory

Ziel

One log format across web, worker and AI service so a log shipper can parse and correlate all three without per-service rules. Today web/worker log pino JSON (time, lower-case level labels such as "info", fixed key whitelist – #28 / PR #45) while the AI service logs ts and upper-case levels.

Akzeptanzkriterien

  • AI service log lines use the same keys as logEvent: time (ISO 8601), level (lower-case label), event, and correlation keys requestId, jobId, companyId where known.
  • The AI service keeps its privacy guarantee: no document content, quotes, names or e-mail addresses in logs (existing tests keep passing; one test asserts the new key set).
  • docs/technical/operations.md documents the shared format once.

Nicht Teil dieser Aufgabe

  • Log shipping, alerting, dashboards.

Betroffene Bereiche

  • services/ai/src/requestflow_ai/ (logging setup), services/ai/tests/, docs/technical/operations.md

Testplan

Kriterium Prüfart
Shared key set + format pytest capturing a log line
No content in logs existing privacy tests + new assertion

Security/Privacy betroffen?

Logs are a data-leak path – keep the whitelist approach; no request bodies.

Found in the security review of PR #45.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Labels

featureNew capabilityreadyDefinition of Ready met – may be claimed

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions