Skip to content

DOC: Clarify framework security boundaries and component reuse - #4465

Open
Richard Lundeen (richlundeen) wants to merge 5 commits into
microsoft:mainfrom
richlundeen:richlundeen-framework-security-boundary
Open

Richard Lundeen (richlundeen) wants to merge 5 commits into
microsoft:mainfrom
richlundeen:richlundeen-framework-security-boundary

Conversation

@richlundeen

@richlundeen Richard Lundeen (richlundeen) commented Nov 20, 2026 •

Copy link
Copy Markdown
Contributor

Description

Clarify the framework's shared-deployment trust model: signed-in users are trusted, and administrator control over server code execution is a soft operational boundary rather than a defense against users.

  • Ask for clear security boundaries enforced the same way everywhere, so nothing fails by accident through inconsistent checks.
  • Allow any user to create shared targets, converters, and scorers using server credentials, including the team's managed identity.
  • Add a rule of thumb to prioritize keeping PyRIT usable for testing unusual situations, while preserving the existing credential-exposure and target-output guidance.
  • Clarify that reuse limits new mechanisms, not the catalog: more named converters and attack techniques (and, to a lesser degree, scenarios and scorers) are welcome, built on the closest existing piece.

Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Copilot AI balanced review requested due to automatic review settings November 20, 2026 06:48

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🟢 Approval recommended

No active review issues were identified for this documentation-only clarification of the shared-deployment trust model.

No open issues 🎉

What changed in this PR

This documentation-only PR clarifies PyRIT’s trust model for shared deployments without changing runtime behavior.

Changes:

  • Documents that authenticated users are trusted to create shared targets, converters, and scorers using server credentials.
  • Clarifies that administrator control over server code execution is an operational boundary, not a defense against trusted users.
  • Preserves owner-only access to private targets and existing guidance on credential exposure and safe handling of target output.
File Description
doc/​code/​framework.md Updates security guidance to clarify shared-deployment trust and private-target boundaries.

Review effort: Balanced

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@richlundeen Richard Lundeen (richlundeen) changed the title DOC: Clarify framework security boundaries DOC: Clarify framework security boundaries and component reuse Nov 20, 2026
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants