Content
92%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A tight, well-structured control skill: exact commands, exact clarification prompts, explicit anti-pattern guardrails, and clear pre-checks. The only weakness is that the preferred MCP-control path lacks a concrete invocation and the mode selection isn't reflected as a placeholder in the start command, leaving minor gaps in actionability.
Suggestions
Show the start command with a mode placeholder (e.g. --mode <mode>, default generic) so non-generic modes from step 2 map visibly onto the CLI invocation.
Add one concrete cue for the MCP path, such as how to discover the available copilot control tools (e.g. their naming pattern) so the preferred path is as actionable as the CLI fallback, while still honoring 'Never invent an MCP tool name'.
Optionally add a post-start verification step (e.g. run minutes copilot status to confirm the session is live) to close the loop after the one-line success report.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~40-line body is lean and assumes Claude's competence: it contains no concept explanations, no padding, and gives exact clarification strings ("Do you want me in this terminal to be your sidekick, or should I open the Minutes Coach HUD?") rather than vague advice, so every token earns its place. | 5 / 5 |
Actionability | The CLI commands ("minutes copilot start --goal '<meeting goal>' --surface tui --mode generic", status/pause/resume/stop) are concrete and executable, but the preferred MCP path is abstract ("call their start control" with no named tool or example call) and the start example hardcodes --mode generic while step 2 defines seven modes — minor gaps that keep it below fully copy-paste-ready coverage. | 4 / 5 |
Workflow Clarity | Each procedure (start, control, output) is a single unambiguous action with an explicit pre-check ("Do not start a second session when minutes copilot status reports one active") and failure handling ("If the provider is unavailable, say that Coach degraded or did not start"), which satisfies the simple-skill exception for a clear, well-sequenced control-plane skill. | 5 / 5 |
Progressive Disclosure | The skill is under 50 lines with no need for external references, and its well-organized sections (Start Coach, Control an active session, Output, Guardrails) with clear headers meet the rubric's simple-skill criterion for a top score; no bundle files exist to reference. | 5 / 5 |
Total | 19 / 20 Passed |