Lean body of concrete cat/python commands with no padding or explanation of concepts Claude already knows; the only prose (steps 3, 5, 7) carries genuinely useful routing guidance, so every token earns its place.

Most steps are fully executable, copy-paste-ready commands with exact paths (cat AGENTS.md, python3 ./.trellis/scripts/get_context.py --mode packages, cat .trellis/spec/frontend/index.md); the one judgment-only step (3) is a reasoning step, not pseudocode, so it stays at the top anchor.

A clearly numbered 1-8 sequence for a read-only prep task; no validation feedback loop is required because the operation is non-destructive, and the mandatory-before-coding framing provides the checkpoint, so it is not capped at 2.

Under 50 lines with no bundle files (references/scripts/assets all absent) and no need for external references; the numbered-step organization is sufficient for the top anchor per the simple-skills note.

Lists multiple concrete actions — 'Discovers and injects project-specific coding guidelines', 'Reads spec indexes, pre-development checklists, and shared thinking guides' — matching the multiple-specific-actions anchor.

Explicitly answers both what (discover/inject guidelines from .trellis/spec/) and when (an explicit 'Use when...' clause with several triggers), matching the both-what-and-when anchor.

Natural triggers a developer would actually say are present: 'starting a new coding task, before writing any code, switching to a different package, or needing to refresh project conventions', giving good coverage rather than jargon.

The .trellis/spec/ niche and pre-development timing give it a clear, distinct trigger surface unlikely to conflict with other skills; not below because the triggers are specific rather than generic.