Conciseness

Every line earns its place: the tool families are compact one-line lists, and non-obvious gotchas ("REST handlers whitelist fields and drop unknown ones", "This runs hybrid BM25 plus vector plus graph-expanded search") are stated without padding. Nothing explains concepts Claude already knows, matching the 'lean and efficient; every token earns its place' anchor rather than 4's 'minor instances of over-explanation'.

Actionability

The quick start fully specifies the common cases with exact tool and parameter names ("memory_save with content (the insight), concepts (comma-separated keywords), files (comma-separated paths)" and "memory_smart_search with query, limit, and targetLayer: 'memory'") — copy-paste-ready for an MCP skill, satisfying the 5 anchor's 'specific examples cover the common cases'. It stays at 5 rather than 4 because the only deferred detail (exact requiredness of other tools' parameters) is explicitly routed to REFERENCE.md rather than left as a gap.

Workflow Clarity

The 4-step workflow is clearly sequenced and includes an explicit verification checkpoint with a recovery loop: "distinguish matched_count: 0 from excluded_by_budget > 0. Expand clipped or omitted IDs with memory_smart_search and expandIds, preserving the same scope filters." This matches the 5 anchor (explicit validation steps plus error-recovery feedback). No destructive/batch cap applies since the workflow is tool selection and calling, not bulk mutation.

Progressive Disclosure

The body is a clear overview with a well-signaled, one-level-deep reference ("Full tool table with parameters and the core-set marking lives in REFERENCE.md, generated from source so it never drifts") and a 'See also' section for sibling skills; the bulk tool table is appropriately split out. No bundle directories (references/, scripts/, assets/) are provided, so scoring follows the body's structure, which matches the 5 anchor's 'content appropriately split; easy navigation'.

Specificity

"Map of every agentmemory MCP tool, what each does, and its parameters" names the domain and three concrete content facets, matching the 'lists several specific actions; minor gaps' anchor. It falls short of 5 because it describes the deliverable (a map) without enumerating any concrete capabilities (e.g., save, search, sessions, knowledge graph), and is clearly above 3's '1-2 concrete actions' since tool coverage, per-tool behavior, and parameters are all specified.

Completeness

Both questions are answered explicitly: what ("Map of every agentmemory MCP tool, what each does, and its parameters") and when ("Use when choosing which memory tool to call, when a tool name or argument is unclear, or when answering what agentmemory can do via MCP") with three concrete trigger phrases. This directly matches the 5 anchor; voice is third person, so no voice penalty applies.

Trigger Term Quality

Natural phrases a user would say are present: "which memory tool to call", "when a tool name or argument is unclear", "what agentmemory can do via MCP", plus keywords like "parameters" and "MCP tool". Not 5 because a few natural synonyms are missing (e.g., "MCP server", "memory function", "API"); not 3 because the phrasing goes beyond bare domain keywords to cover multiple real trigger situations.

Distinctiveness Conflict Risk

The "agentmemory MCP tool" namespace is a clear niche with distinct triggers and explicit MCP scoping, so conflict risk is minimal. The only conceivable overlap is with sibling skills (agentmemory-rest-api, agentmemory-config), but those cover different surfaces, keeping this at the 'clear niche' 5 anchor rather than 4.