Surface multiscan run warnings in campaign summaries
Issue #248 also calls for run warnings to be reflected in the multiscan campaign summary, similar to failures.\n\nPR #810 persists run warnings in results.jsonl and replays them on resume, but intentionally leaves campaign summary accounting unchanged. A repository can therefore complete with persisted operational warnings while the top-level summary still counts it as an ordinary completed scan.\n\nFollow-up scope:\n- decide how operational run warnings should appear in MultiscanResult / CLI campaign summary without conflating them with incomplete coverage;\n- preserve existing completed status semantics for successful scans;\n- add summary-level regression coverage for warned-but-complete repositories and resumed receipts.
Source: openai/codex-security