Conciseness

Every section delivers xberg-specific facts Claude would not know (default `--format json` vs `text`, layout concurrency limits of 1/2, `--output-dir` must exist, path-key matching rules) with zero padding or explanation of known concepts. This matches anchor 5 — lean, assumes competence, every token earns its place — and shows none of the over-explanation that would drop it to 4.

Actionability

Nearly every section contains copy-paste-ready commands: `xberg batch *.pdf`, a jq post-processing pipeline, a complete `--file-configs` JSON example, `mkdir -p` + `--output-dir`, a TOML config, and a runnable Python `extract_batch` snippet covering the common cases. The only nit is `await` shown at top level, which is the standard convention for illustrating async APIs and keeps it at anchor 5 rather than revealing a real gap.

Workflow Clarity

For a batch operation the rubric caps workflow_clarity at 3 without concrete validation, and this skill's checkpoint is implicit: "Inspect results for partial content and surfaced errors rather than relying on the process exit code alone" gives direction but no executable validation command (e.g., a jq filter for per-file errors) and no retry/fix loop for failed files. This fits anchor 3 (checkpoints missing or implicit); it does not reach 4 because the verification step lacks a concrete command, and is above 2 because error handling is at least explicitly addressed.

Progressive Disclosure

Sections are well-organized and the bulk flag detail is correctly deferred ("See `references/cli-reference.md` for the full `batch` flag set"), but the bundle contains no `references/` directory at all — that pointer is dangling, and the API references are ambiguously located "in the sibling `xberg` skill". Per the guideline to score against the actual bundle structure, this matches anchor 3 (references present but not clearly signaled/verifiable) rather than 4's 'references mostly clear'.

Specificity

The description lists several concrete capabilities — "shared config, bounded parallelism, per-file overrides, and error recovery" plus the "`batch` command", "`--file-configs`", "`--max-concurrent`", and "output layout" — which matches the anchor for several specific actions with minor gaps. It is not 5 because coverage is incomplete (no file formats, no mention of the structured array output), and not 3 because it goes well beyond naming 1-2 actions.

Completeness

Both parts are present: the what ("shared config, bounded parallelism, per-file overrides, and error recovery. Covers the `batch` command...") and an explicit when ("Use when extracting from many files at once"). It matches anchor 4 rather than 5 because the 'when' relies on a single trigger clause without additional concrete trigger phrases or variations.

Trigger Term Quality

"Use when extracting from many files at once" is a natural phrase a user would say, and "batch" plus the concrete flag names give good keyword coverage. It falls short of 5 because common variations like "bulk", "folder/directory of documents", and file-type keywords are missing, but it clearly exceeds the 3 anchor's 'missing common variations' level.

Distinctiveness Conflict Risk

"many files at once" and the `batch` command clearly carve out a niche distinct from single-file extraction, but as a sibling of the broader `xberg` extraction skill there is minor overlap risk — a user asking to 'extract these PDFs' could plausibly need either skill. This fits anchor 4 (mostly distinct, minor overlap with closely related skills), not 5's minimal-conflict profile.