Conciseness

The body is a lean command reference with no explanations of concepts Claude already knows; only minor trimming is needed — 'Each argument is separate — do NOT quote subcommands together' is repeated in both the Organization Management and Skill Management sections, and 'Getting Started' re-states the auth command already listed above. Anchor 4, not 5.

Actionability

Highly concrete throughout: exact grep commands, the full auth-login.js subcommand list, and a complete skillify flag reference. Minor gaps keep it below anchor 5: the auth script path is an unresolved placeholder ('$PLUGIN_ROOT/bundle/commands/auth-login.js ... or check the session context', later '<AUTH_CMD>'), and the availability of the 'hivemind' CLI on PATH is never stated.

Workflow Clarity

'How to Search' is a well-sequenced 4-step read path (index → summaries → raw JSONL → grep) with an explicit 'Do NOT jump straight to reading raw JSONL' guardrail, but the skill includes destructive and batch operations (remove <user-id>, skillify unpull, embeddings uninstall --prune, skillify pull --force) with no validation or verification steps such as confirming a user ID via 'members' or running --dry-run before destructive runs — the rubric caps this at 3.

Progressive Disclosure

No bundle files exist, and the single-file body is well-sectioned with clear headers and a memory-structure tree that make each area scannable. Minor gaps: ~100 lines of CLI reference (the skillify flag list alone is ~25 lines) that could live in a separate reference file, and the auth-login.js pointer reaches outside the skill without clear signaling. Anchor 4, not 5 — not an ideal split with well-signaled one-level-deep references.

Specificity

Names the domain ('Global team and org memory powered by Activeloop') and one concrete action ('ALWAYS check BOTH built-in memory AND Hivemind memory when recalling information'), but omits the skill's other capabilities (searching, org management, skillify, embeddings) — matching the '1-2 concrete actions, not comprehensive' anchor rather than the 'several specific actions' anchor.

Completeness

Both 'what' ('Global team and org memory powered by Activeloop') and 'when' ('ALWAYS check ... when recalling information') are explicitly present, so it clears anchor 3's 'when missing or weakly implied', but the 'what' names no concrete actions and the 'when' covers only recall — not the explicit, concrete trigger phrases of anchor 5.

Trigger Term Quality

'memory' and 'recalling information' are natural user terms, but common synonyms ('remember', 'shared knowledge', 'past sessions') are absent and product jargon ('Activeloop', 'Hivemind') carries the description, fitting the 'some relevant keywords but missing common variations' anchor.

Distinctiveness Conflict Risk

The named product ('Hivemind memory', 'Activeloop') carves a distinct niche and explicitly differentiates from 'built-in memory', but the broad recall trigger fires on every memory-recall request, creating minor overlap with the built-in memory skill — anchor 4, not 5.