What
The datum-schema skill defines the canonical type taxonomy and field schema for the _b00t_ ecosystem. Supported datum types include: skill (skill definitions), cli (CLI tool wrappers), mcp (MCP server definitions), agent (sub-agent definitions), role (role definitions with skills/compliance/entanglements), gate (governance gates), schema (data protocol schemas), and hardware (hardware capability descriptions).
Required [b00t] fields for all datums: name (unique kebab-case identifier), type (one of the datum types above), and hint (human-readable description, ≤120 chars). Skill-specific fields under [b00t.skill] include depends_on (prerequisite skills), unlocks (capabilities unlocked when loaded), description, tags, and metadata. The tail-map format (# b00t:map v1) requires: summary (≤120 chars), tags, tier (sm0l/ch0nky/frontier), cmds, and complexity (1-10).
When to Use
Load this skill when authoring new datums, validating existing ones, or understanding the b00t type system. It is a prerequisite for datum-authoring.
How
- Identify the datum type that matches your resource (skill, cli, mcp, agent, etc.).
- Fill required
[b00t]fields: name, type, hint. - If creating a skill, use
[b00t.skill]with depends_on, unlocks, description, tags. - Append the
# b00t:map v1tail-map with summary, tags, tier, cmds, complexity. - Verify: no
[b00t.version](use[b00t.schema]instead), no[b00t.type_tags](use tail-map tags).