SpeedTree Subsystem Audit

Use the Claude Code command as the canonical workflow instead of maintaining a duplicated snapshot.

Load the Canonical Command

1. Read .claude/commands/audit-speedtree/SKILL.md completely.
2. Read .claude/commands/_audit-common.md and .claude/commands/_audit-severity.md completely before auditing.
3. Follow the canonical command's settled-fact premises, Phase 1 setup (default plus recon feature lanes and the deep-only corpus harness), the six risk-ordered dimensions with the per-finding Dimension field, and report format.

Adapt Claude Conventions to Codex

• Treat the user's text after the skill request as the command arguments described by the canonical skill's argument-hint.
• Resolve each /audit-<name> reference to that audit's SKILL.md under .claude/commands/, and route Skyrim-plus BSTreeNode trees to audit-nif, distant-LOD and real-tree-geometry design to audit-exterior, and NIFAL single-boundary material findings to audit-nifal.
• Prefer the repository's codebase-memory graph tools for TREE wiring and import-path tracing; use text search for tag numbers, .spt literals, guard test names, and graph gaps.
• Interpret Claude Task-agent instructions as Codex sub-agent delegation only when delegation is available and allowed; the canonical architecture is single-pass, so run all dimensions inline in its risk order.
• Preserve the canonical output path and do not create GitHub issues; publishing remains a separate audit-publish action.

Do not copy the canonical command body into this skill. The Claude command is the single source of truth.