Add qualitative content assessment to the agent skill

[mindsocket]

Context

The agent skill (skills/structured-content/) currently covers structural validation and tooling operations. A second comment on #2 called out a richer use case: using rules to evaluate content quality — e.g. is an opportunity framed in customer voice? Is it in the problem space?

The rules capability exists (from #16), but neither bundled schema (strict_ost, general) has any rules. This means agents can't currently use schemas show to retrieve qualitative guidance.

What's needed

• Add example qualitative/descriptive best-practice rules to bundled schemas (strict_ost and/or general) — both to demonstrate the capability and to give agents something useful to query via schemas show
• Update the skill (SKILL.md or a new reference doc) to explain how to use schemas show to retrieve qualitative rules and apply them when reviewing content
• Primary source for rule content: docs/torres-alignment-research.md

Goal

Close the loop on the vision from #2: "the schema evolves from pure structural validation into a richer specification that encodes both 'is this file structurally valid?' and 'is this content well-formed?'"

Acceptance criteria

• At least one bundled schema has meaningful qualitative rules (e.g. opportunity framing, problem vs solution space)
• schemas show --space <alias> displays those rules to an agent
• SKILL.md (or a reference doc) explains the qualitative assessment workflow