Content
81%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.
Highly actionable content with precise MCP tool invocations, a well-sequenced workflow, and strong safety gating around irreversible deletions. The main weakness is redundancy — the intro restates the description and key cautions are repeated multiple times, which inflates token cost without adding guidance.
Suggestions
Remove the opening paragraph that restates the description and workflow, letting the Prerequisites/Core Principles sections carry that framing.
State 'Configs cannot be archived' once (e.g., in Core Principles) and remove the three repeat mentions in Step 2, the retire section, and What NOT to Do.
Consolidate the repeated confirm:true / 'suggest turning targeting off first' guidance into a single confirmation convention defined before the workflow steps.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense and operational with no filler concept explanations, but it could be tightened: the opening paragraph restates the description, 'Configs cannot be archived' appears four times, and the confirm:/suggest-targeting-off guidance repeats across sections. | 3 / 5 |
Actionability | For an instruction-only skill it gives fully executable guidance: exact MCP tool names, resource values, and parameters ('get-agent-configs' with configKey and section: "health", 'update-agent-config' with resource: "targeting", on: false, confirm: true, maintainerId, modelConfigKey, defaultVariationKey), covering update, retire, delete, restore, and verify. | 5 / 5 |
Workflow Clarity | A clear three-step sequence (assess health/state, make the change, verify) with an explicit verification step, error-recovery guidance ('Flag any issues or rollback if needed'; retry-by-verifying instead of trusting the echo), and destructive operations gated behind explicit user confirmation. | 5 / 5 |
Progressive Disclosure | A single well-sectioned SKILL.md (no bundle files exist) with a clearly signaled external docs link and a Related Skills section; structure is good and content is appropriately placed, but at ~95 lines with no progressive-disclosure split it sits just below the top anchor. | 4 / 5 |
Total | 17 / 20 Passed |