feat(adapters): Hermes agent adapter

[aktasbatuhan]

Goal

Add an adapter so watchmen captures Hermes agent sessions alongside the other supported agents. One corpus across tools.

Adapter contract

Per src/watchmen/adapters/init.py, a new adapter is a module exposing:

• NAME: str
• discover() -> Iterable[FileEntry] (walk the install path, yield sessions; return [] when the Hermes agent isn't installed)
• scan(entry) -> (session_dict, prompts, tool_calls)

It registers in the ADAPTERS tuple. claude_code / codex are the cleanest references for a JSONL-on-disk source; cursor.py is the reference for a SQLite key-value store.

Open questions (resolve before coding)

• What exactly is the target here (confirm which Hermes agent / runtime), and where does it persist sessions, in what format?
• Per-turn token usage / cost: available and reliable? If not, leave at 0 (no price-table guessing, like the cursor adapter).
• Tool-call + outcome signal we can map to is_error / friction signatures (feat: friction ledger — recurring mistakes grouped by error signature (#110) #110)?
• Skill detection path (SKILL.md heuristic, like the non-Claude adapters)?

Acceptance criteria

• Confirm the target agent and document its on-disk session format
• Implement discover() + scan() against synthetic fixtures
• Register in ADAPTERS and add to the README agent list
• Unit tests covering the conversation shape(s), tool calls, and skill detection
• Full suite green, no regressions

Note: target + storage format are unconfirmed; first task is to pin both down before writing the parser.