Conciseness

The body is operational throughout — commands, exit-code branches, and gating instructions with almost no explanation of concepts Claude already knows. Minor trimmable material keeps it below the score-5 'lean and efficient' anchor: the Constraints section restates the frontmatter description nearly verbatim, and the template before/after examples in Phase 4 are somewhat long, though they encode genuinely non-obvious instructions; it is clearly above score 3, where padding would be more than incidental.

Actionability

Guidance is mostly executable: the orchestrator invocation with --pr/--channel/--action flags (verified against scripts/flutter_cp.dart, whose interface, exit codes 0/1/2, and SUCCESS:AUTOMATED:/SUCCESS:MANUAL: markers all match), `git status --porcelain`, `gh issue view <ISSUE_NUMBER> --json title,body`, and a complete `gh pr edit` command with title format and label. Below the score-5 anchor because a few placeholders are never resolved — `<DART_EXECUTABLE>` has no guidance on how to determine it, and `--body "<FINAL_TEMPLATE>"` gives no handling for multi-line bodies in shell.

Workflow Clarity

The four-phase workflow has an explicit sequence, validation checkpoints (clean-repo check via `git status --porcelain`, exit-code-driven branching at each script run), a full error-recovery feedback loop (exit 2 → analyze conflicts → resolve or hand to user → `git add` → `--action continue`), two explicit gating steps ('MUST STOP CALLING TOOLS AND ASK' for the channel choice and for user approval of the drafted template), and a fallback for `gh pr edit` GraphQL failures. This matches the score-5 anchor: clear sequence, explicit validation, and feedback loops for error recovery.

Progressive Disclosure

Structure is good and appropriately split: the 460-line automation logic correctly lives in scripts/flutter_cp.dart (a real bundle file, correctly referenced and accurately described), the template lives in the repo's .github directory (one level deep), and the body stays a lean orchestration guide with clear phase headers. Below the score-5 anchor due to minor organization gaps: the template is referenced via an awkward inline-code-plus-markdown-link hybrid ('[PULL_REQUEST_CP_TEMPLATE](../../../.github/PR_TEMPLATE/PULL_REQUEST_CP_TEMPLATE.md) relative to this skill') and the script path appears only inside bash commands with no explicit listing or description of the bundle contents.

Specificity

The description names one concrete action — "land a formal cherry-pick of a merged PR for the flutter/flutter repo stable or beta channel" — but does not enumerate the several specific capabilities the skill performs (creating the CP PR, filling the CP template, resolving conflicts, adding the cp: review label). This matches the anchor "Names domain and 1-2 concrete actions, but not comprehensive"; it is below score 4 because it does not list several specific actions, and above score 2 because the action named is concrete and domain-precise.

Completeness

Both questions are explicitly answered with concrete triggers: the what ("land a formal cherry-pick of a merged PR for the flutter/flutter repo stable or beta channel") and the when ("Only use for flutter/flutter landed pull requests. Only use when the cherry pick request is into 'stable', 'beta' or a branch that has the format with flutter-<major>.<minor>-candidate.0"). This matches the score-5 anchor — explicit 'what' and 'when' with concrete trigger phrases — rather than score 4, where the 'when' would be less specific.

Trigger Term Quality

Good natural-keyword coverage: "cherry-pick"/"cherry pick", "merged PR", "landed pull requests", "stable", "beta", and the concrete candidate-branch format. It falls short of the score-5 anchor because common synonyms users actually say are missing — notably the "CP" abbreviation (used by Flutter contributors, e.g. "CP PR 187045") and "backport" — but it is well above score 3, where only some relevant keywords are present with common variations missing.

Distinctiveness Conflict Risk

The description carves out a clear niche: cherry-picking already-merged PRs in the flutter/flutter repository only, targeted at stable, beta, or flutter-<major>.<minor>-candidate.<hotfix> branches. These are distinct triggers with minimal overlap risk against any other skill, matching the score-5 anchor rather than score 4 ("minor overlap risk").