fix(converter): de-duplicate skill descriptions in command wrappers

[ak2k]

Summary

Each skill was showing up twice in interactive Claude Code's /context skills list:

├ ce-compound-refresh: ~170 tokens          ← SKILL.md description
├ ce-lite:ce-compound-refresh: ~60 tokens   ← commands/<name>.md description

For ce-lite's 88 command wrappers, that's ~6k tokens of duplicated description text.

Root cause

generate_commands.py was emitting a shortened SKILL.md description: into every commands/<name>.md frontmatter. Claude Code counts each as a separate skill entry: bare-named for SKILL.md, plugin-namespaced for the command wrapper.

Fix

Command-wrapper description: is now just the skill name itself (description: "ce-plan" instead of description: "Create structured plans..."). Satisfies validate.py's non-empty-description check while collapsing each duplicate entry to ~5 tokens.

The SKILL.md description is still the canonical routing signal. The command wrapper exists only for /<plugin>:<name> dispatch reliability (see generate_commands.py module docstring for why command wrappers exist at all).

shorten_description + MAX_DESCRIPTION_LEN are dropped — no longer used.

How this was discovered

Measured interactively in #27's A/B run. The +7.9k ce-lite-side skills bump (ce-lite 12.5k vs upstream 4.6k) was the smoking gun — 88 skills vs upstream's 29 explained part of it, but the +60-token-per-command duplication explained the rest.

Expected impact

After merge + publish-dist regen:

• ce-lite Skills slot: 12.5k → ~6.5k
• Total savings vs upstream: −105.2k → ~−111k

PR #27's README A/B table will need a refresh in a follow-up.

Test plan

• nix flake check green (treefmt + actionlint + 161 pytest)
• Re-ran full pipeline locally against pinned upstream tag; validate.py --upstream byte-equivalent round-trip still passes
• After merge, trigger publish-dist and re-measure with isolated CLAUDE_CONFIG_DIRs