Clarify "unshipped Learning output style" wording in learning-output-style plugin docs

The learning-output-style plugin describes itself as mimicking the "unshipped Learning output style", but Learning is a shipped built-in output style — listed under Built-in output styles at https://code.claude.com/docs/en/output-styles.

The wording appears in 4 files (verified on main, 2026-08-18):

• plugins/learning-output-style/README.md (3×)
• plugins/README.md (plugins index table)
• plugins/learning-output-style/hooks-handlers/session-start.sh (2×)
• plugins/learning-output-style/.claude-plugin/plugin.json (description)

Changelog timeline: Learning shipped in v1.0.81, was deprecated in v2.0.30 and un-deprecated in v2.0.32 (and /output-style has since been folded into /config).

The plugin isn't identical to the built-in style (generic TODO markers vs the built-in's TODO(human); it also merges in Explanatory-style ★ Insight behavior), so one reading is that "unshipped" refers to an earlier internal prototype of the Learning prompt that never shipped as-is — the README's phrase "differs from the original unshipped Learning output style" supports that. Alternatively the wording dates from the v2.0.30–v2.0.32 deprecation window and is now stale.

Could you clarify what "unshipped" is meant to convey? Happy to open a small PR updating the wording in all four files once the intended meaning is confirmed (e.g. "based on an early prototype of the Learning output style", or simply "the Learning output style").