perf(profiles): allow lightweight profile listing

[0disoft]

What does this PR do?

Adds opt-in lightweight profile listing so callers that only need profile names and directories do not pay every per-profile detail cost.

list_profiles() still returns the same full ProfileInfo data by default. New keyword-only flags let hot paths skip specific expensive details:

• include_model skips config.yaml parsing
• include_distribution skips distribution.yaml parsing
• include_description skips profile.yaml parsing
• include_gateway_status skips gateway PID/runtime-state checks
• include_skill_count skips recursive skills/**/SKILL.md counting
• include_alias skips wrapper alias scanning
• include_env skips .env existence checks

This keeps detailed surfaces such as hermes profile list and /api/profiles unchanged, while cheaper routing/sync paths use minimal profile records.

Related Issue

N/A. I searched open issues/PRs for profile list skills count performance and did not find a direct duplicate.

Type of Change

• 🐛 Bug fix (non-breaking change that fixes an issue)
• ✨ New feature (non-breaking change that adds functionality)
• 🔒 Security fix
• 📝 Documentation update
• ✅ Tests (adding or improving test coverage)
• ♻️ Refactor (no behavior change)
• 🎯 New skill (bundled or hub)

Changes Made

• hermes_cli/profiles.py
  • Adds keyword-only include flags to list_profiles() with all defaults preserving current behavior.
  • Skips optional I/O and recursive scans when callers request lightweight profile records.
• Updated lightweight call sites that only need name/path or a narrow subset of fields:
  • update bundled-skill sync
  • session search cross-profile lookup
  • dashboard profile-session aggregation
  • gateway process discovery and gateway list
  • doctor profile checks
  • uninstall named-profile discovery
  • kanban decomposition profile roster
  • profile description listing
  • Honcho profile sync/status helpers
  • kanban dashboard profile roster
• Tests:
  • Adds a list_profiles() test proving expensive helpers are not called when detail flags are disabled.
  • Updates a session-search test double to accept the new keyword-only call shape.

How to Test

1. Focused profile listing tests:

   uv run --with pytest --with pytest-timeout --with pytest-xdist python -m pytest tests/hermes_cli/test_profiles.py::TestListProfiles -q

   Result: 5 passed

2. Distribution metadata compatibility:

   uv run --with pytest --with pytest-timeout --with pytest-xdist python -m pytest tests/hermes_cli/test_profile_distribution.py -q

   Result: 69 passed

3. Kanban roster behavior:

   uv run --with pytest --with pytest-timeout --with pytest-xdist python -m pytest tests/hermes_cli/test_kanban_decompose.py -q

   Result: 9 passed

4. Dashboard profile/session paths:

   uv run --with pytest --with pytest-timeout --with pytest-xdist python -m pytest tests/hermes_cli/test_web_server.py -k "profiles_sessions or profiles_list" -q

   Result: 4 passed, 311 deselected

5. Session search:

   uv run --with pytest --with pytest-timeout --with pytest-xdist python -m pytest tests/tools/test_session_search.py -q

   Result: 47 passed

6. Update profile skill sync:

   uv run --with pytest --with pytest-timeout --with pytest-xdist python -m pytest tests/hermes_cli/test_cmd_update.py::TestCmdUpdateProfileSkillSync -q

   Result: 2 passed

7. Syntax and whitespace checks:

   python -m py_compile hermes_cli/profiles.py hermes_cli/main.py hermes_cli/gateway.py hermes_cli/doctor.py hermes_cli/uninstall.py hermes_cli/kanban_decompose.py hermes_cli/profile_describer.py hermes_cli/web_server.py tools/session_search_tool.py plugins/memory/honcho/cli.py plugins/kanban/dashboard/plugin_api.py tests/hermes_cli/test_profiles.py tests/tools/test_session_search.py
   git diff --check

   Result: passed.

I also tried a broader direct pytest run for tests/hermes_cli/test_profiles.py tests/hermes_cli/test_profile_distribution.py tests/hermes_cli/test_kanban_decompose.py; the new/related tests passed, but 5 existing profile tests failed on this Windows direct-pytest environment because of existing cp949 default decoding and Windows chmod mode behavior. I did not change those unrelated tests in this PR.

Checklist

Code

• I've read the Contributing Guide
• My commit messages follow Conventional Commits (fix(scope):, feat(scope):, etc.)
• I searched for existing PRs to make sure this isn't a duplicate
• My PR contains only changes related to this fix/feature (no unrelated commits)
• I've run pytest tests/ -q and all tests pass
• I've added tests for my changes (required for bug fixes, strongly encouraged for features)
• I've tested on my platform: Windows 11

Documentation & Housekeeping

• I've updated relevant documentation (README, docs/, docstrings) — N/A, internal API behavior documented in code
• I've updated cli-config.yaml.example if I added/changed config keys — N/A
• I've updated CONTRIBUTING.md or AGENTS.md if I changed architecture or workflows — N/A
• I've considered cross-platform impact (Windows, macOS) per the compatibility guide
• I've updated tool descriptions/schemas if I changed tool behavior — N/A

For New Skills

N/A.

Screenshots / Logs

N/A. This is an internal profile-listing performance path change.

AI assistance: this PR was prepared with Codex assistance; I reviewed the diff and ran the verification listed above.