Conciseness

The body is dense and imperative with no concept-teaching padding, but the 18-entry Error Handling catalog restates roughly six conditions already covered inline (e.g., "gh CLI not authenticated", "Label missing from repository", "No issues match filters") and "If no results match the filters…" appears verbatim in both tracker branches. Efficient overall with trimming available — anchor 4, not 5; the systematic duplication keeps it above 3 only because the rest is lean.

Actionability

Fully executable throughout: copy-paste-ready gh commands with exact flags and JSON fields ("gh issue list --label \"<label1>,<label2>\" --state open --limit 10 --json number,title,labels,createdAt,author"), exact JQL, exact display formats, a verbatim drift-detector agent prompt, and concrete file-detection paths. Matches the top anchor; the generic 'use the Atlassian MCP server' phrasing is inherent to MCP integration and doesn't leave a real gap.

Workflow Clarity

Steps 1–7 are clearly sequenced with explicit validation (git remote check, "gh auth status", MCP availability probe), a gated Step 6.5 with skip conditions, severity-based feedback loops (none/minor → proceed, moderate/major → ask user), and failure fallback ("If the drift detector itself fails … Do not block on a tool failure"). Batch label creation is check-then-act and flagged idempotent, so the destructive/batch cap does not apply — top anchor.

Progressive Disclosure

The greenfield procedure is properly externalized to a real, well-signaled one-level-deep reference ("Read ${CLAUDE_PLUGIN_ROOT}/skills/arn-code-pick-issue/references/greenfield-backlog-resolution.md for the full detection chain, resolution workflow, and sub-feature handling"), and cross-references to platform-labels.md and ensure-config.md are clearly signaled. However, the ~287-line body inlines a large error catalog that partially duplicates both the steps and the reference file — good structure with minor organization gaps, anchor 4 rather than the lean-overview anchor 5.

Specificity

Quotes several concrete actions — "browse issues filtered by Arness labels, select one, and route it to the appropriate Arness pipeline skill for implementation" plus "local-first dependency resolution from a greenfield feature backlog" — matching the anchor for several specific actions with minor gaps. Not 5 because the assessment/rejection loop and spec routing outcomes are not covered; not 3 because far more than 1-2 actions are named.

Completeness

Explicitly answers both: what ("browse issues filtered by Arness labels, select one, and route it to the appropriate Arness pipeline skill"; "local-first dependency resolution") and when ("should be used when the user says … or wants to browse issues"), with concrete trigger phrases and a negative boundary ("Do NOT use this for creating new issues — use /arn-code-create-issue"). Clearly the top anchor, not 4, since the 'when' is fully explicit with concrete triggers.

Trigger Term Quality

Comprehensive natural phrasings a user would actually say: "pick issue", "grab issue", "what should I work on", "next issue", "what's unblocked", "pick from feature tracker", plus the exact skill name. Matches the comprehensive-coverage anchor; no common variation is obviously missing, so not 4.

Distinctiveness Conflict Risk

Clear niche — issue selection and routing within the Arness pipeline — with explicit disambiguation against the sibling skill ("Do NOT use this for creating new issues"). Generic phrases like "show issues" carry at most minimal overlap risk, matching the clear-niche/minimal-conflict anchor rather than the minor-overlap anchor at 4.