[DRAFT] FEAT Normalize simulated conversation prompt sources

[hannahwestra25]

Description

Work on #2494 highlighted a gap in the shared factory API: benchmark- or scenario-owned adversarial prompts may already exist as SeedPrompt objects, but with_simulated_conversation() only had a path-shaped input. That forced callers to keep prompt files coupled to the shared technique catalog or add one-off adaptation around the benchmark.

This PR pulls that general capability out as an independent change based directly on main. The preferred adversarial_chat_system_prompt input now accepts either an inline SeedPrompt or an explicit pathlib.Path. The factory resolves either form immediately and stores the canonical inline prompt, so #2494 and similar callers can own prompt content without moving prompt packaging into executors or changing persistence semantics. Plain strings are intentionally not accepted by the preferred input because non-path prompt APIs use strings for inline text; legacy string paths remain supported through adversarial_chat_system_prompt_path.

The old path-named parameter remains a silent compatibility alias for this first step, and direct or persisted legacy path-backed SeedSimulatedConversation records still execute normally. The other simulated-conversation *_system_prompt_path inputs stay unchanged here because generalizing them requires separate model, persistence, and execution compatibility work. This PR does not depend on #2494 or change AdversarialBenchmark itself.

Tests and Documentation

• 397 focused and adjacent unit tests passed
• Ruff check/format, ty, async-suffix, compile, and diff checks passed
• Jupytext paired sources and round-trip checks passed
• Updated the attack-techniques documentation pair