Content
85%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.
This is a well-structured, instruction-only orientation skill with explicit routing, drift-validation checkpoints, and lean prose that assumes Claude's competence. The main lever for improvement is tightening the few repeated repository-filter caveats and adding one concrete worked query example to reach fully copy-paste-ready actionability.
Suggestions
Consolidate the repeated 'apply metadata_filters={"repository": ...}' and 'never run an unscoped coding-session query' warnings into a single canonical rule referenced from each route, to trim repeated tokens.
Add one short concrete worked example of a topic-discovery query (e.g. an example read_note/search_note call with realistic arguments) to push actionability from concrete-but-instructional toward copy-paste-ready.
Optionally factor the query parameter table (type/status/repository/timeframe per route) into a compact reference block so each route section can stay purely about decision logic.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and assumes Claude's competence (it never explains what Basic Memory, MCP, or read_note are), with only minor repetition of the repository-filter caveats across routes that could be tightened slightly; it sits above the score-3 anchor but is not a 5 because those warnings recur and could be consolidated. | 4 / 5 |
Actionability | It gives concrete, executable guidance (call read_note with project=<configured primaryProject>, apply metadata_filters={"repository": ...}, query type=task status=active) rather than vague direction; per the code_vs_instruction note, absence of code in an instruction-only skill is not penalized, but it stops short of fully copy-paste-ready examples. | 4 / 5 |
Workflow Clarity | The process is explicitly sequenced (Resolve Configuration -> Choose One Recall Route -> Check Current State -> Present and Continue) with clear validation checkpoints: drift comparisons (same/different SHA, branch, repo), explicit error-reporting branches ('no credible match', 'setup incomplete'), and a feedback structure for ambiguous cases. | 5 / 5 |
Progressive Disclosure | The skill is a single well-organized SKILL.md with clear section headers and no nested or buried references (the only bundle file is assets/icon.svg); per the simple-skill scoring note, well-organized single-file content with no need for external references scores 5. | 5 / 5 |
Total | 18 / 20 Passed |