Agent Skills: tomllm-format

What

The TOML+LLM format defines conventions for enriching valid TOML with comment annotations optimized for LLM consumption. .tomllm (machine-oriented) and .tomllmd (documentation-heavy) files extend plain .toml with three comment conventions: # @tribal: / # 🤓 for non-obvious knowledge, design rationale, and gotchas; # @example: for usage examples inline with the datum; and # b00t:map v1 as a structured tail-map footer (last ≤10 lines of the file).

The tail-map contains: summary (one-line description, ≤120 chars), tags (comma-separated keywords), tier (sm0l/ch0nky/frontier), cmds (comma-separated relevant commands), and complexity (1-10 integer). The critical rule: the tail-map MUST be the last ≤10 lines of the file — nothing follows it.

When to Use

Use .tomllmd for documentation-heavy datums like PRDs and schemas. Use .tomllm for machine-oriented configuration datums. Use plain .toml as the fallback. Load this skill when authoring any new b00t datum.

How

1. Write valid TOML as normal.
2. Add # 🤓 / # @tribal: comments for non-obvious design rationale.
3. Add # @example: comments for usage examples.
4. Append # b00t:map v1 as the footer with summary, tags, tier, cmds, complexity.
5. Ensure nothing follows the tail-map — it is a footer, not a middle section.