Content
75%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 well-structured advisory skill body: terse, metric-anchored, with concrete triggers, output contracts, and proper use of one-level-deep reference files. Weaknesses are minor redundancy and several cross-references to skills and directories that are not present in this bundle.
Suggestions
Delete the 'Keywords' section (it duplicates the description's trigger terms) and the second copy of the stage-adaptive horizons in the Tree of Thought section, pointing back to the first instead.
Remove or qualify the references to 'culture-architect/', 'board-deck-builder/', and '../agent-protocol/SKILL.md' — none exist in this bundle, so they are dead ends; if they are sibling skills, say so explicitly.
Add one example invocation with sample input and output for each script in Quick Start so the commands are demonstrably runnable without opening the script files.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and imperative (tables, terse bullets, no explaining what a CEO or OKR is), but there are avoidable duplicates: the 'Keywords' section restates the description's triggers, the stage-adaptive horizons appear twice (Vision & Strategy and Tree of Thought sections), and references/board_governance_investor_relations.md is pointed to twice. These are minor trims, not the padded over-explanation of the 3 anchor. | 4 / 5 |
Actionability | Concrete and executable throughout: copy-paste script commands in Quick Start, specific numeric targets (>70% goals hit, >12 months runway, <2x burn multiple, >40 NPS), a capital-allocation priority order, and a request→artifact mapping table. Not 5 because the scripts lack example invocations with inputs/outputs, and the '[INVOKE:role|question]' syntax is used without any explanation of its semantics. | 4 / 5 |
Workflow Clarity | Multi-step processes are clearly sequenced with evaluation criteria: the strategic planning cycle (annual/quarterly/monthly), Tree of Thought (generate 3 paths → evaluate upside/downside/reversibility/second-order effects → pick), proactive condition→action triggers, and an output pipeline with verification stages (self-verify, peer-verify, critic pre-screen, confidence tags). Not 5 because the Internal Quality Loop's actual steps are delegated to an external skill file rather than stated inline. | 4 / 5 |
Progressive Disclosure | Good structure: the body stays an overview, all three references/*.md files exist, are one level deep, are signaled in context, and are catalogued in a Resources section with one-line descriptions; scripts are surfaced in Quick Start. Not 5 because the body points to 'culture-architect/', 'board-deck-builder/', and '../agent-protocol/SKILL.md', none of which exist in this bundle — navigation dead ends for a reader. | 4 / 5 |
Total | 16 / 20 Passed |