Conciseness

The ~30-line body is lean with no explanation of concepts Claude already knows and no padding, e.g., "Treat code-notify as an optional local dependency managed outside this repository" and "If missing, tell the user the integration is unavailable in the current shell and continue without blocking the main task". It misses 5 because the verify-before-relying guidance is repeated across three sections ("Verify the installed command, wrapper, hook, or config before using it", workflow step 2, and "Mention that code-notify is available only after verifying") and meta-advice like "Prefer concise setup notes over tool-specific prose duplication" could be trimmed, matching 'Efficient; minor instances of over-explanation that could be trimmed'.

Actionability

The guidance is directive and conditionally complete ("use it for end-of-task or attention-needed notifications only", fallback behavior in step 4), but every operational step is an abstract intent with no concrete mechanism: "Detect whether the current tool already has a notification hook or wrapper configured" and "Look for a local code-notify command, alias, script, or documented install path" give no commands, config file paths, or example invocations. This matches 'Some concrete guidance but incomplete; missing key details' — above 2 because the workflow does specify what to do in each outcome, below 4 because nothing in the body is directly executable.

Workflow Clarity

The 4-step workflow is a clear, ordered sequence with an explicit presence-verification gate and fallback branch: detect existing hook, "Look for a local code-notify command... before assuming invocation details", use it if present, otherwise "tell the user the integration is unavailable... and continue without blocking". This matches 'Clear sequence with most checkpoints present; minor validation gaps' — it does not reach 5 because the detection steps have no explicit method (no command or file to check), and exceeds 3 because the verify-before-use checkpoint and failure path are explicit rather than implicit.

Progressive Disclosure

The skill is under 50 lines, self-contained, and needs no external references; the body is organized into five clear, well-scoped sections (Goal, Use This Skill When, Cross-Platform Rules, Workflow, Output Guidance) with no inlined material that belongs in separate files. Per the rubric's simple-skill guideline, this earns 5 with just well-organized sections and no bundle files to misplace.

Specificity

The description names the domain and a couple of concrete actions — "trigger a local completion notification through code-notify", "configuring cross-tool notification behavior for Codex, Cursor, Claude, Gemini", "Verify the installed command or existing local integration" — but the actions stay at a high level (no concrete operations comparable to "extract text, fill forms"), matching the anchor 'Names domain and 1-2 concrete actions, but not comprehensive'. It does not reach 4 because the actions are abstract intents rather than specific operations, and is above 2 because the domain and actions are clearly identified.

Completeness

Both elements are present: an explicit "Use when a task should trigger a local completion notification through code-notify, or when configuring cross-tool notification behavior" clause, and a 'what' conveyed through the same capabilities plus the action "Verify the installed command or existing local integration before relying on it". It does not reach 5 because the 'what' is not stated as a crisp standalone capability declaration — it is woven into the 'when' clause — matching 'Has both what and when; when could be more explicit or specific' only in the sense that the what/when split is blurred; it clearly exceeds 3 since the 'Use when' trigger guidance is explicit.

Trigger Term Quality

Strong natural triggers are present: "completion notification", "code-notify", and the explicit tool names "Codex, Cursor, Claude, Gemini" that a user would naturally mention. It falls short of 5 because common phrasing variants users would say — "notify me when it's done", "alert", "ping me" — are missing, matching 'Good keyword coverage; a few natural terms missing'. It is above 3 because coverage goes beyond a single generic keyword and includes multiple distinct, natural trigger terms.

Distinctiveness Conflict Risk

The niche is well defined — local completion notifications via the named "code-notify" command, scoped to "Codex, Cursor, Claude, Gemini, or similar agent workflows" — giving distinct triggers with only minor overlap risk against generic notification-setup skills. It matches 'Mostly distinct; minor overlap risk with closely related skills' rather than 5 because "similar agent workflows" and general notification language leave slight overlap with broader notification/config skills.