feat: record telemetry span outcomes

[codeforester]

Fixes #392

Summary

• set OpenTelemetry span status from the normalized lifecycle outcome
• record the original exception on failed invocations when available
• cover successful and failed telemetry paths

Branch maintenance

Refs #426. Targets main.

The branch already contained main at a576cc279739eae5e4cfc33ffab2a7fb56de24de. No new commit was needed.

Current-head validation

At f32811439ab15bf72b8ee1d77e467852c78ea8dd: uv lock freshness and baseline, runtime, strict typing, style, and contracts passed locally with all declared extras. Runtime result: 609 passed, 1 warning, 262 subtests passed in 7.65s.

Hosted checks: 7/7 required checks passed; 0 checks pending; 0 unsuccessful checks at 2026-10-04T14:19:40.760978+00:00. See the PR Checks tab and #426 for subsequent results.

[codeforester]

Resolved and revalidated.

• The original failures were benchmark-contract outliers (Unix persistence-enabled p95 and WSL cold-import p95); the full WSL test suite itself reported 606 passed.
• Reran the failed hosted jobs: Unix benchmark and WSL runtime/benchmark both passed.
• Synced this branch with current main after security: guard delimited spreadsheet formulas #404 landed, then reran the complete Tests workflow: all 23 jobs passed.
• Local validation: 606 passed; Ruff, formatting, and strict mypy all pass.

The PR is now clean/mergeable and remains open for review.