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.
An exceptionally actionable reference with explicit workflows, validation checkpoints, and error-recovery loops for a complex toolchain domain. Its weaknesses are repetition and over-nested prose in the Critical Directives (hurting conciseness) and references to bundle files that are not present alongside the SKILL.md.
Suggestions
State the projectPath rule once in the Critical Directives and have the later notes ("Which project the tools act on", the install note) reference it instead of restating it, cutting a significant block of repeated text.
Move the multi-project repository rules and the install/refresh contract (each a long nested blockquote) into a reference file, keeping only a one-line pointer in SKILL.md.
Ship the referenced bundle files (references/*.md including references/analysis-tools/, scripts/mps_dump.py, scripts/examples/) with the skill, or remove references to files that are not part of the bundle, so the signaled navigation paths actually resolve.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense with genuinely non-obvious domain knowledge (tool routing, ordering chains, exact error strings), but the projectPath rule is stated three times (the Critical Directive, the "Which project the tools act on" note, and the install note), and several directives are heavily nested walls of parenthetical sub-clauses (e.g. the scaffold-chain and multi-project notes). Mostly efficient but noticeably tighten-able — anchor 3, not 4. | 3 / 5 |
Actionability | Guidance is fully executable: exact tool names with the stable mps_mcp_ prefix, exact parameter keys (nodeReference, conceptRefs, parameters.code), executable script invocations with real arguments, literal error strings ("Unable to determine the target project…", LANGUAGE_NOT_BUILT), and named recovery calls. Specific examples cover the common cases. | 5 / 5 |
Workflow Clarity | Multi-step processes are explicitly sequenced with validation checkpoints and recovery loops: the numbered structure-edit → clean MAKE → verify descriptors → scaffold chain, the 5-step session-init workflow, "Validate frequently" with mps_mcp_check_root_node_problems after every mutation, rebuild-after-compiled-aspect-change rules, and failure recovery from rejection messages and EXPLAIN_TYPE. Matches the anchor-5 feedback-loop pattern. | 5 / 5 |
Progressive Disclosure | The overview uses well-signaled, one-level-deep references ("Open references/finding-things.md for…", "See references/analysis-tools/type-queries.md") plus a companion-skills index table — good structure. However, none of the referenced files exist in the bundle (no references/, scripts/, or assets/ directories), so navigation is broken as presented, keeping this below anchor 5. | 4 / 5 |
Total | 17 / 20 Passed |