Add saga orchestrator with compensating rollback

[JE-Chen]

Workflow batch (from multi-agent web research; verified absent — long-listed, never built). Full layers + tests + EN/Zh v52 docs + README.

Feature (utils/saga, pure-stdlib)

• Saga records a compensating action per step; on any step failure, runs the completed steps' compensations in LIFO order — the durable-transaction primitive AC_try (single-block try/catch/finally) couldn't provide. SagaResult carries ok/completed/compensated/failed_step/error. Compensation is best-effort (a failing undo is logged, rollback continues).
• Forward actions/compensations are plain callables (or, via run_saga, JSON action lists run through the executor), so it's fully unit-tested with no side effects.
• Executor AC_run_saga (steps as list or JSON string); MCP ac_run_saga; Builder under Flow.

Verification

• 7 tests pass (all-succeed no-compensation, failure→LIFO compensation, steps-without-compensation skipped, best-effort compensation continues past a failing undo, executor round-trip, wiring, facade); ruff clean; radon no CC≥C; bandit clean; PySide6-free.
• Rebased onto dev after JSONPath (Add JSONPath querying over parsed JSON #259) merge.

[codacy-production]

Up to standards ✅

🟢 Issues 0 issues

Results:
0 new issues

View in Codacy

🟢 Metrics 32 complexity · 0 duplication

Metric	Results
Complexity	32
Duplication	0

View in Codacy

_{NEW Get contextual insights on your PRs based on Codacy's metrics, along with PR and Jira context, without leaving GitHub. Enable AI reviewer
TIP This summary will be updated as you push new changes.}

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud