feat: improve skill scores for 5 skills + add Tessl review workflow

[rohan-tessl]

Hey @lamm-mit 👋

I ran your skills through tessl skill review at work and found some targeted improvements.

Here's the full before/after:

Skill	Before	After	Change
alphafold	0%	85%	+85%
boltz	0%	89%	+89%
chai	0%	89%	+89%
disease-research	25%	79%	+54%
drug-research	31%	93%	+62%

This PR intentionally caps changes at 5 skills to keep it reviewable. The included GitHub Action workflow (see below) will surface Tessl feedback on future SKILL.md changes automatically — so the rest of the 300+ skill library can improve incrementally.

What changed

alphafold, boltz, chai (0% → 85–89%):

• Added missing YAML frontmatter (name, description, metadata) — these skills had no frontmatter at all, which caused deterministic validation failure
• Added actionable "Use when…" description clauses
• Converted body text to third-person voice

disease-research (25% → 79%):

• Merged duplicate frontmatter blocks into one valid block
• Moved source into metadata to avoid unknown-key warnings
• Trimmed body from 638 lines → ~60 lines; moved heavy tool reference tables, code examples, and report templates to references/tool-reference.md (per CONTRIBUTING.md pattern)
• Added structured workflow table for the 10 research dimensions

drug-research (31% → 93%):

• Merged duplicate frontmatter blocks into one valid block
• Moved source into metadata
• Trimmed body from ~1592 lines → ~120 lines; moved detailed tool chains, disambiguation workflows, evidence grading, and templates to references/tool-reference.md
• Structured the workflow overview and report sections as concise tables

All changes follow skills/CONTRIBUTING.md: YAML frontmatter with kebab-case names, quoted descriptions, and heavy content in references/ directories.

🤖 Included: Tessl Skill Review GitHub Action

This PR also adds .github/workflows/skill-review.yml so future PRs that touch SKILL.md files get automatic review feedback:

• What runs: On PRs that change **/SKILL.md, the workflow runs tesslio/skill-review and posts one comment with Tessl scores and feedback (updated on new pushes).
• Zero extra accounts: Contributors do not need a Tessl login — only the repo's default GITHUB_TOKEN is used to post comments.
• Non-blocking by default: Unless you add fail-threshold, the check is feedback-only — no surprise red CI.
• Not a build replacement: This is review automation for skill markdown, not a substitute for your own pipelines.
• Optional gate: Add with: fail-threshold: 70 later if you want PRs to fail on low scores.
• Why only 5 skills here: This PR caps manual optimization so it stays reviewable; after merge, every PR that touches SKILL.md gets automatic review comments, so the rest of the library improves incrementally.

---

Honest disclosure — I work at @tesslio where we build tooling around skills like these. Not a pitch — just saw room for improvement and wanted to contribute.

Want to self-improve your skills? Just point your agent (Claude Code, Codex, etc.) at this Tessl guide and ask it to optimize your skill. Ping me — @rohan-tessl — if you hit any snags.

Thanks in advance 🙏