Clarify live PR triage behavior in docs

[auscaster]

The public docs should explain that support-triage responds to issues first and only then escalates bounded work into issue-to-pr.

Acceptance criteria:

• docs/flows.md reflects the triage-first behavior
• the hosted Sourcey docs still build
• no unrelated repo mutations occur

Retry marker: rerun after workflow and runx contract fixes.

[github-actions]

Thanks for the clear report and acceptance criteria. This looks like a bounded docs-only change: update docs/flows.md to reflect the triage-first behavior, keep the hosted Sourcey docs build passing, and avoid unrelated repo mutations. We should route this into a single PR-sized docs fix.

[auscaster]

Rationale: Approved to continue one bounded docs-only issue-triage run for this issue.
Approved By: Kam
Applies To: issue-triage.*
Objective Fingerprint: 21e360bc15136f50
Expires After: 2026-04-27T00:00:00Z
Invariant: Keep the work docs-only.
Invariant: Keep the work inside nilstate/aster.
Invariant: No unrelated repo mutations.
Note: Prefer a single bounded PR if the lane escalates past triage.