Skip to content

Context

Context represents why access is needed. In Alien Giraffe this is a small set of real fields on the request itself, plus the audit trail those fields end up in — not a live sync with HR or ITSM systems.

A request’s justification is submitted alongside its scope and constraints in the same payload — see Policies for how all five components combine.

Every request template and request requires:

  • business_justification — free-text explanation of why the access is needed.
  • description — a short summary of the request.

And optionally carries:

  • external_ticket_id / external_ticket_url — a reference to wherever the work is tracked (a Jira, ServiceNow, or Linear ticket, for example). This is a link an approver can click; Alien Giraffe does not synchronize with the ticketing system, create tickets automatically, or read status back from it.
  • source (manual | ticket | ai) — how the request originated.

An approver reviewing a request in Approve And Manage Requests sees business_justification, description, risk_level, and the ticket link together, and approves or denies based on that context.

Every meaningful event tied to a request is written to the audit_logs table with actor identity, action type, resource type/id, level, IP address, and structured metadata — including:

  • request approval, denial, and revocation
  • tunnel creation, deletion, and credential use
  • explorer query execution
  • tunneled query execution

If a Fluentd sidecar is enabled, runtime logs from env-enzo are additionally enriched with request/environment metadata and can be fanned out to PostgreSQL, Slack, or S3 — see Observability & Audit and Auditing with FluentD. This is the mechanism for “why did this access happen” after the fact; it complements the up-front justification fields, it doesn’t replace them.

There is no live integration with HR systems (Workday, BambooHR), ITSM platforms (ServiceNow, Jira Service Management), or GRC/compliance tools (Drata, Vanta, OneTrust) today. external_ticket_url is the extent of ticket-system integration — a link field a requester fills in, not a synced record.

  • Policies — how justification combines with the other four components
  • Subjects — the identity every audit entry is attributed to
  • Resources — the datasource/dataset a justification is written for
  • Constraints — the approval decision made using this context
  • Channels — where audited query activity actually happens