feat: add first-party skills, ponytail lock, and full lint suite

- add ponytail skill set to skills-lock.json (14 locked skills total)
- add first-party medusa skill under skills/, layered on install
- make test now runs ruff, shellcheck, and markdownlint-cli2
- CI installs the lint tools before make test in both workflows
- MIT license, funding buttons, thanks and donation sections in README
- .markdownlint-cli2.jsonc config; ignore lint caches in .gitignore

skills/medusa/SKILL.md

@@ -0,0 +1,32 @@
+---
+name: medusa
+description: >
+  Run the MEDUSA AI security scanner (medusa-security on PyPI) against a
+  repository, working tree, or secret stores. Use when the user asks for a
+  security scan, supply-chain vetting of a repo before cloning, leaked-secret
+  detection in AI chat or shell history, or invokes /medusa. Defensive scanning
+  only.
+---
+
+MEDUSA is a defensive AI-first security scanner (40,000+ patterns: CVEs, prompt injection, MCP tool poisoning, leaked secrets).
+
+## Setup
+
+Check availability with `medusa --version`.
+If missing, install it with `pipx install medusa-security` (fallback: `pip install --user medusa-security`).
+
+## Commands
+
+| Task | Command |
+| --- | --- |
+| Scan the current project | `medusa scan .` |
+| Vet a remote repo BEFORE cloning (supply chain, poisoned `.claude/` hooks, permissions, skills) | `medusa scan --git <url-or-org/repo>` |
+| Find leaked API keys in AI chat + shell history | `medusa secrets scan` |
+| Scan one exported file | `medusa secrets scan --path <file>` |
+
+## Rules
+
+- Run scans read-only; never pass `--reveal` and never run `medusa secrets purge` unless the user explicitly asks for it.
+- Summarize findings by severity with file/line references; do not paste secret values into the conversation, even redacted ones.
+- Treat findings as signals for review, not verdicts; confirm true positives in the code before proposing fixes.
+- Vet third-party agent repos (`medusa scan --git`) before recommending the user clone or install them.