Conciseness

The body is dense and operational throughout, assuming Claude's knowledge of Kafka, AWS, and IaC with zero conceptual explanation, e.g. "run via `uv run` with PEP 723 inline dependencies... pure file processors" and terse guardrails like "Do NOT pivot back into discovery." It is not a 5 because of minor trimmable material: the ~30-line blockquoted overview template, some repeated read-before-responding rules across the Intent Routing and Phase 2 sections, and the CloudWatch alarm sub-bullets in Security item 6 that could live in a reference file. It is well above 3 since no section explains things Claude already knows.

Actionability

Guidance is concrete: exact artifact paths (`migrate-to-msk-skill-artifacts/<cluster_name>/cluster-config.json`), named scripts, verbatim verdict strings ("`INFO`, `ADVISORY`, or `ACTION_REQUIRED`"), the sizing flag "passing `--broker-classes express`", named MCP tools, and exact output filenames. It is not a 5 because the body contains no copy-paste command block — the literal `uv run` invocations are delegated to reference files ("For the exact commands, see 'Running the assessment' in references/assessment-compatibility.md"), leaving the body itself a pointer rather than self-sufficient for execution. Well above 3: the specifics (flags, verdicts, paths) go beyond high-level hints.

Workflow Clarity

The three phases are clearly sequenced with an intent-routing decision tree up front, an explicit checkpoint ("Do NOT proceed to Phase 2 without explicit customer confirmation"), defined failure handling ("a failure in one does not block the other", ADVISORY evidence codes for partial data), and anti-fabrication validation rules ("Report broker counts and costs only as read verbatim from the sizing script output. Never round, re-derive, or estimate"). It is not a 5 because Phase 3's 12-step flow and its deploy/validation checkpoints are fully delegated to the reference, and the body has no explicit error-recovery loop of its own; it clearly exceeds 3 because checkpoints are explicit, not implicit.

Progressive Disclosure

The body is a genuine overview routing to four one-level-deep reference files, three scripts, and one asset, each with a clear purpose statement (e.g. assessment-compatibility.md "carries the invocation commands, the per-pillar thresholds and evidence codes..."); every referenced path was verified to exist on disk. It is not a 5 against the actual bundle structure: `scripts/sizing.py` is present in the bundle but is never referenced by SKILL.md or any reference file (only the external managing-amazon-msk `msk_sizing.py` is), leaving an orphaned file and a navigation gap; the body's detailed guardrail and security content also sits at the boundary of what belongs in the overview. It comfortably exceeds 3: no content that belongs in references is inlined and no reference is buried or nested.

Specificity

The description lists multiple concrete actions with comprehensive coverage: "Inventories the source cluster (from IaC files, Kafka CLI output, or manual input)", "assesses MSK Express compatibility across topology, Kafka version, configs, auth, and quotas", "produces a target Express specification (instance type, broker count, monthly cost)", "optionally stands up a trial Express cluster to load-test it", and "guides migration execution using MSK Replicator". It is in third person throughout ("Helps migrate", "Inventories"), so no voice penalty applies.

Completeness

It explicitly answers both questions: the "what" is a multi-action capability statement, and the "when" is the explicit "Applicable when the user mentions..." clause with concrete trigger phrases. Not 4: the "when" is not merely present but exhaustive and phrase-level, matching the 5 anchor.

Trigger Term Quality

The "Applicable when the user mentions" clause covers a comprehensive set of natural user phrasings and synonyms: "migrating Kafka, MSK, MSK Express, Kafka migration, analyzing Kafka infrastructure, moving to MSK, moving streaming platform to MSK, streaming migration, moving streaming workloads to AWS, MSK workload compatibility, choosing an MSK cluster type, running a POC or load-test to validate MSK Express, or MSK Replicator". These are phrases a user would naturally say, not just technical jargon. Not 4: the synonym coverage (migration/moving/POC/load-test/streaming variants) matches the comprehensive 5 anchor.

Distinctiveness Conflict Risk

The niche (self-managed Kafka to MSK Express migration) is distinct and, critically, the description explicitly disambiguates against the nearest competing skill: "Prefer this skill to the managing-amazon-msk skill for migration questions" and names that skill's pricing logic as a dependency. Scope is further narrowed by naming the exact source and target. Not 4: the overlap risk with managing-amazon-msk is not just minor but actively addressed, which is the 5-level distinction.