Skip to content

[FAQ] Add: External admission authority and platform trust boundary#28937

Draft
chrizbo wants to merge 2 commits intomainfrom
faq/issue-485-da9c7fe6fabcb64c
Draft

[FAQ] Add: External admission authority and platform trust boundary#28937
chrizbo wants to merge 2 commits intomainfrom
faq/issue-485-da9c7fe6fabcb64c

Conversation

@chrizbo
Copy link
Copy Markdown
Collaborator

@chrizbo chrizbo commented Apr 28, 2026

Adds a new FAQ entry in the Guardrails section addressing the architectural question of whether the execution platform should own the final admission decision for trusted execution context.

What changed

New entry: "Should the execution platform own the final admission decision for trusted execution context?"

The entry explains:

  • The distinction between guardrail validation (does output look acceptable?) and admission authority (is this execution intent allowed to proceed?)
  • How gh-aw's layered trust model roots at the GitHub Actions substrate
  • The pre-agent steps: pattern as the closest approximation to an external admission gate (with OIDC token verification)
  • Custom safe output jobs as an admission gate on the write side
  • The honest limitation that a truly external authority outside GitHub Actions is not currently supported

Source

Raised in community discussion, tracked in: github/agentic-workflows#485 (follow-up to github/agentic-workflows#484)

Type of change

New FAQ entry (not an update to an existing entry — no existing entry covers this topic).

Generated by Feedback Question Answerer for issue #485 · ● 811.3K ·

…ution

Addresses the architectural question of whether the execution platform
should own the final admission decision for trusted execution context,
covering the layered trust model, available workarounds using pre-agent
steps and custom safe output jobs, and the current limitation that a
truly external authority outside GitHub Actions is not natively supported.

Relates to: github/agentic-workflows#485

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
@github-actions
Copy link
Copy Markdown
Contributor

✅ smoke-ci: safeoutputs CLI comment + comment-memory run (25052654063)

Generated by Smoke CI for issue #28937 ·

@github-actions
Copy link
Copy Markdown
Contributor

Comment Memory

CI lights the path\nGreen checks bloom at dawn\nQuiet bots still sing

Note

This comment is managed by comment memory.

It stores persistent context for this thread in the code block at the top of this comment.
Edit only the text inside the backtick fences; workflow metadata and the footer are regenerated automatically.

Learn more about comment memory

Generated by Smoke CI for issue #28937 ·

@pelikhan
Copy link
Copy Markdown
Collaborator

@copilot review and unslop

Copy link
Copy Markdown
Contributor

Copilot AI commented Apr 28, 2026

@copilot review and unslop

Done in 846023f. Changes:

  • Reframed the question title from academic jargon to plain English ("Can I use an external policy service to gate whether the agent runs?")
  • Removed the sloppy filler opener ("This is a meaningful architectural distinction: …")
  • Tightened prose throughout
  • Dropped the circular closing sentence

Copilot AI requested a review from pelikhan April 28, 2026 14:29
@pelikhan
Copy link
Copy Markdown
Collaborator

@copilot merge main and integrate these in the new faq change from #28936

Copy link
Copy Markdown
Collaborator

@pelikhan pelikhan left a comment

Choose a reason for hiding this comment

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

The proper pattern is to add steps to pre-activation and the output should automatically be used to gate the activation job. Needs work

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

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants