Skip to main content
addly

Every permission needs a reason, scope, and owner.

This page describes the security standard expected for future addly apps. It does not turn concepts into audited products and claims no certification that has not been obtained.

Guiding principle

Request the minimum, explain every access, respect host permissions, and document uninstall behavior.

Four layers, four responsibilities.

A Forge app does not become secure merely by being close to the host. Every boundary remains named and testable.

1

Host product

Jira or Confluence identity, permissions, and authorized context.

2

Application

Scopes, access controls, input validation, and secrets.

3

Optional services

Authentication, encryption, logging, and subprocessors.

4

Operations

Access reviews, vulnerability management, and incident response.

Intended permissions, never presented as a real manifest.

ConceptReadWriteExpected control
Readiness for JiraRead the issue contextWrite only app-owned statesAlignment with the published manifest
Portfolio InsightsRead selected projectsWrite only view preferencesAlignment with the published manifest
Page SummaryRead the current pageWrite only macro configurationAlignment with the published manifest
Context LinkerRead the current pageRead Jira only for accessible projectsAlignment with the published manifest

Report a vulnerability

The future channel must be private, acknowledged quickly, and separate from functional support. No SLA is invented here; the production procedure still needs publication.

View the escalation path

Evidence to attach

  • 01Forge manifest and scope justification.
  • 02Secret inventory and rotation.
  • 03Threat model and trust boundaries.
  • 04Deletion and uninstall procedure.
  • 05Publishable security-fix history.

Useful security is app-specific.

Open the documentation for the selected concept to connect permissions, configuration, and data to the behavior that is actually demonstrated.