Conciseness

The body is mostly tight tables and command blocks, but there is real duplication: the 'Quick Reference' table repeats commands detailed in Core Commands, and complexity guidance appears twice (the Complexity Tiers table and the 'Choosing Complexity' bullets). The env-var table also hardcodes a dated model version ('claude-opus-4-5-20251101') that will go stale, matching the 'mostly efficient but could be tightened' anchor rather than the minor-trim profile of level 4.

Actionability

Every section gives copy-paste-ready bash commands with comments, full option tables for spec_runner.py and run.py, and worked end-to-end examples ('Complete Feature Development', 'Quick Bug Fix') that cover the common cases. This matches the fully-executable, copy-paste-ready anchor; not level 4 because even edge flows (pausing, file-based control, debugging) come with exact commands and file paths.

Workflow Clarity

The main workflow is clearly sequenced with checkpoints before the destructive merge step — review changes (step 3) and test in the worktree (step 4) precede '--merge' (step 5), and the QA loop is described as an explicit validate→fix→repeat cycle. It falls short of level 5 because the 'Batch Processing' example runs multiple specs back-to-back with no validation or status check between builds, leaving a minor validation gap.

Progressive Disclosure

The body is well-sectioned with clear headers, but roughly 290 lines of CLI detail are inlined monolithically while the bundle's references/CLI-USAGE.md — a directly overlapping CLI usage guide — and references/CLAUDE.md are never linked or mentioned anywhere. This matches the anchor 'content that should be separate is inline' with references effectively unsignaled, rather than level 4's 'references mostly clear'.

Specificity

The description names the domain ('Auto-Claude CLI command reference and usage patterns') and a few actions ('running specs, managing builds, checking status'), matching the anchor for a domain plus 1-2 concrete actions without comprehensive coverage. It never names the actual commands or entry points (spec_runner.py, run.py, validate_spec.py), so it falls short of the 'several specific actions' of level 4.

Completeness

It explicitly answers both questions: the 'what' ('Auto-Claude CLI command reference and usage patterns') and a concrete 'when' ('Use when running specs, managing builds, checking status, or using CLI commands for autonomous coding tasks'). This matches the level-5 anchor of a clear what paired with explicit trigger phrases; it is not level 4 because the 'when' clause is specific and multi-trigger rather than merely serviceable.

Trigger Term Quality

Triggers like 'running specs', 'managing builds', and 'checking status' are phrases users of this framework would naturally say, giving some relevant keywords. However, common variations and concrete artifacts are missing — no mention of 'spec', 'build', 'worktree', 'merge', or the command names themselves — matching the 'missing common variations or synonyms' anchor rather than the good coverage of level 4.

Distinctiveness Conflict Risk

The named niche (Auto-Claude CLI) provides some distinctiveness, but the trigger phrases overlap heavily with the sibling skills the body itself lists — 'running specs' vs auto-claude-spec, 'managing builds' vs auto-claude-build, workspace management vs auto-claude-workspace. This matches the anchor 'somewhat specific but could still overlap with similar skills' rather than the minor-overlap profile of level 4.