feat(issue-triage): enable native issue-intent on supported safe outputs

Add hyphenated `issue-intent: true` to every configured safe-output tool
that supports it under gh-aw v0.83.0. Inspected all configured outputs
against the v0.83.0 schema:

- add-labels: supported -> issue-intent enabled (native rationale/
  confidence/suggest on suggested labels).
- add-comment: not an issue-intent-capable output type in v0.83.0 -> left
  unchanged.

No safe-output tools were added or removed. Recompiled the issue-triage
workflow with `gh aw compile`; the generated handler config now carries
"add_labels":{...,"issue_intent":true}. frontmatter_hash changes as an
intentional result of this config edit; body_hash is unchanged.

Validated with `gh aw compile` and `gh aw validate` (0 errors, 0 warnings),
idempotent on re-compile.

Refs github/plan-track-agentic-toolkit#519

Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Copilot-Session: 628d4172-49f6-4d5c-905c-a33276de3aee

Author: alondahari

.github/workflows/issue-triage.lock.yml

@@ -45,6 +45,7 @@ safe-outputs:
     client-id: ${{ secrets.CLI_TRIAGE_APP_CLIENT_ID }}
     private-key: ${{ secrets.CLI_TRIAGE_APP_PRIVATE_KEY }}
   add-labels:
+    issue-intent: true
     max: 3
     allowed:
       - bug