feat(skills): add 6 phoenixlucky skills (4 bundled + 2 optional)

[phoenixlucky]

What & Why

Add 6 skills by phoenixlucky, adapted to the hermes SKILL.md format.

Bundled (skills/productivity/) — broadly useful analysis & discipline frameworks:

• business-data-analyst — structured business analysis: metric definitions, funnel/cohort/retention breakdowns, root-cause diagnosis, market research, and strategy/competition/marketing/operations/finance models.
• weiliaozi — five-lens structured analysis framework (Essence → Conditions → Gains-Losses → Sequence → Opponent), bilingual EN/中文.
• financial-analyst — investing analysis: cycle positioning → drivers → constraints → gains/losses → rhythm → game theory.
• zerotoken — token-efficient assistant discipline: task-mode budgeting, precise-prompt template, progressive reading, Wei Liaozi coding principles, plus reusable encoding helper scripts for Windows/Chinese environments.

Optional (optional-skills/) — useful but not universally needed:

• communication/moon-lovers — romantic chat roleplay replies from character profiles (8 profiles, 白月光 tone).
• health/family-lovers — self-healing companion for family-of-origin awareness and safe attachment habits (companion tool, not a substitute for professional care).

How to test

• hermes -s weiliaozi "分析：某公司是否应该进入下沉市场"
• hermes skills browse — optional skills appear as "official"
• hermes skills install moon-lovers

Platforms

• Content verified on Windows (PowerShell); skills are platform-agnostic (linux/macos/windows).

Licensing note

All six skills are original works by phoenixlucky. weiliaozi, financial-analyst, family-lovers are MIT; business-data-analyst and moon-lovers are contributed under MIT; zerotoken was previously GPL-3.0 and is relicensed to MIT by the author for this contribution (the original repo remains GPL-3.0).

[Enough1122]

AI code review — automated review for reference, author can ignore or act on any point.

feat(skills): add 6 phoenixlucky skills

• All six SKILL.md description fields exceed the 60-character limit in the skill authoring standards (AGENTS.md "Skill authoring standards" Terminal tool #1). Examples: optional-skills/communication/moon-lovers/SKILL.md line 9 (~440 chars), skills/productivity/business-data-analyst/SKILL.md (~500 chars), optional-skills/health/family-lovers/SKILL.md (~230 chars), skills/productivity/weiliaozi/SKILL.md (~400 chars). These bloat skill listings and dilute the model's attention when many skills load. Consider trimming each to a single ≤60-char capability sentence and moving details into the body.
• No tests shipped at tests/skills/test_<name>_skill.py (standard some cleanups #7: skill tests live there, stdlib + pytest only, no live network). zerotoken bundles scripts/test_safe_io.py inside the skill directory, which the repo convention does not pick up in the CI skill-test lane. Consider relocating/adding tests under tests/skills/.
• optional-skills/health/family-lovers/SKILL.md positions itself as therapeutic ("在陪伴性对话中获得疗愈", inner-child / 原生家庭 trauma work). Recommend adding an explicit disclaimer that it is not a substitute for professional mental-health support, mirroring the 18+ labeling used in the moon-lovers reference profiles — the framing can otherwise be read as clinical guidance.
• Minor: several reference files end with trailing blank lines (chen-xiang-profile.md, hong-hai-er-profile.md, hua-qian-gu-profile.md, jin-chan-zi-profile.md) — trim before merge.

[Enough1122]

Content reviewed: all six frontmatters are well-formed (name/description/version/author/license/platforms/metadata.hermes), all referenced references/ and scripts/ files exist, no skill-name collisions, and the target categories (skills/productivity, optional-skills/communication, optional-skills/health) already exist. One housekeeping gap: the PR's licensing note is prominent (including the zerotoken GPL-3.0→MIT relicensing), yet only 3 of the 6 skills ship a LICENSE file — financial-analyst, weiliaozi, and family-lovers have one, while business-data-analyst, zerotoken, and moon-lovers do not. Since the note explicitly calls out licensing terms for each skill, adding the missing LICENSE files (and for zerotoken, the relicensing note) would make the contribution self-consistent and avoid any ambiguity for downstream redistribution.