perf(e2e): reuse one authorized Telegram Desktop session per Mantis run (#127835)

* refactor(mantis): reuse authorized desktop captures

* fix(mantis): budget desktop authorization failures

* chore(mantis): bound desktop proof retries

* fix(e2e): drop unused recorder failure fact type export

* fix(ci): route Mantis desktop teardown through the recorder wrapper

Cleanup invoked the internal recorder executable as mantis-sut, which is
deliberately kept out of the docker group and cannot read the
recorder-owned session file; teardown therefore failed and blocked
safe_to_release. The cleanup step already runs as the recorder user, so
call the public wrapper whose exec shim cds into the session root.

* fix(e2e): make recorder failure fact lane-readable; document v2 lifecycle

The Mantis workflow runs the recorder as the desktop user while the lane
reads the authorization-failure fact as mantis-sut; 0600 made that read
fail EACCES and silently disabled the two-attempt retry budget. Write the
fact 0644 — the 0770 attempt directory bounds visibility.

Update the mantis doc's recorder section for the v2 session lifecycle:
required --session handle with healthy-session reuse, capture-only stop,
and teardown owning authorization termination and lease release.

.github/codex/prompts/mantis-telegram-desktop-proof.md

@@ -139,7 +139,9 @@ behavior, call `block`; do not call `finish` and describe the block only in pros
 Inspect `mantis-lane-facts.json`, every returned event/request, the inspection
 PNG, final PNG, and cropped GIF. Confirm the evaluated message is fully visible
 near the bottom and the recording covers the behavior—not only its final state.
-Iterate as needed; all attempts remain recorded.
+Iteration is allowed, but if `start` reports `desktop-unavailable`, record that
+fact and use `block`; never retry that lane. Two non-advancing repeats of the
+same failing step mean classify and stop, not retry. All attempts remain recorded.
 
 If you design a novel working scenario worth reusing, optionally write
 `MANTIS_OUTPUT_DIR/recipe-suggestion.md` with its trigger, exact commands, and

.github/workflows/mantis-telegram-desktop-proof.yml

@@ -273,7 +273,7 @@ jobs:
     needs: resolve_request
     if: needs.resolve_request.outputs.should_run == 'true' && needs.resolve_request.outputs.publish_artifact_name == ''
     runs-on: blacksmith-16vcpu-ubuntu-2404
-    timeout-minutes: 360
+    timeout-minutes: 120
     environment: qa-live-shared
     outputs:
       comparison_status: ${{ steps.inspect.outputs.comparison_status }}
@@ -934,6 +934,11 @@ jobs:
                 abort --lane "$lane" || result=1
             fi
           done
+          # Teardown goes through the public wrapper: this step already runs as the
+          # recorder user, and the internal exec shim cds into the session root where
+          # the recorder-owned session file lives. mantis-sut can do neither.
+          /usr/local/bin/openclaw-telegram-desktop-recorder \
+            teardown --session desktop-recorder.json || result=1
           if sudo test -f "$lock"; then
             echo "Mantis harness lock remained after cleanup." >&2
             result=1