Conciseness

Lines like "Pass the prompt as one quoted argument" and "Never start the interactive UI or `kimi login`" carry only product-specific facts Claude cannot know (flags, `KIMI_CODE_HOME`, DSH escalation) with no padding and no explanation of concepts Claude already knows. The final Example paragraph is a concrete scenario, not filler, so every token earns its place.

Actionability

Fully executable guidance throughout: `command -v kimi`, `kimi -p "<prompt>" --output-format stream-json`, and the DSH escalation recipe (`sandbox_permissions: danger-full-access` with a justification naming session storage). The closing example covers the common case (approved implementation) end to end.

Workflow Clarity

A clear numbered sequence (set working directory and timeout, check availability, run one prompt) with explicit failure handling ("Treat an error event or nonzero exit as failure"), a retry-once feedback loop for sandbox denials, stop-and-ask on auth/config problems, and post-run diff inspection as validation. Not below 5 because error-recovery checkpoints are explicit rather than implicit.

Progressive Disclosure

A single-purpose skill under 50 lines with two well-organized sections (Run, Authentication And Permissions), no bundle files present, and no dangling references. Per the simple-skill guideline, well-organized sections with no need for external references warrant a 5.

Specificity

"Run Kimi Code non-interactively for coding tasks" names the tool and one concrete action (non-interactive running) but does not enumerate the skill's actual capabilities (parse JSONL output, extract the final result, manage permissions/escalation). This matches anchor 3 ("1-2 concrete actions, not comprehensive") rather than anchor 4's "several specific actions".

Completeness

Both parts are explicit: what ("Run Kimi Code non-interactively for coding tasks") and when ("Use when the user asks to delegate work to Kimi Code"). The single narrow when-clause is less comprehensive than anchor 5's multiple concrete trigger phrases, but far above anchor 3 where the when is missing or only weakly implied.

Trigger Term Quality

"delegate work to Kimi Code" plus the product name itself are natural phrases a user would say, giving good keyword coverage. Common variations ("ask Kimi to…", "run this with Kimi") are missing, keeping it below anchor 5's comprehensive synonym coverage but above anchor 3's partial coverage.

Distinctiveness Conflict Risk

The description names a specific third-party tool with a tool-specific trigger ("delegate work to Kimi Code"), creating a clear niche with minimal conflict risk against generic coding or delegation skills.