Conciseness

The body is dense and operational with no basic-concept explanations, matching 'Efficient; minor instances of over-explanation that could be trimmed' — e.g., the novelty check is described in both Phase 3 and Phase 4, and the Phase 6 report template includes illustrative sample rows. It is lean enough to sit above anchor 3 but has small redundancies that keep it from 5.

Actionability

Fully executable guidance: copy-paste-ready Codex MCP prompt blocks with model/config specified, exact CLI commands (e.g., 'python3 "$ARIS_INFRA/tools/research_wiki.py" add_edge research-wiki/ --from ... --type inspired_by --evidence ...'), numeric constants with override syntax, a GPU deployment layout, and a complete report template — matching 'Fully executable; copy-paste ready code or commands'.

Workflow Clarity

Seven clearly sequenced phases with explicit validation checkpoints and feedback loops: novelty quick-check in Phase 3, devil's-advocate review in Phase 4, pilot success metrics defined upfront, a hard timeout kill rule with partial-result collection, re-ranking on empirical evidence, explicit skip conditions for Phases 0/5/7, and a STOP-and-ask guard for overly broad directions — matching the top anchor.

Progressive Disclosure

No bundle files exist; the single SKILL.md is well-sectioned (Overview, Constants, phased Workflow, Key Rules, Composing) with clearly signaled, one-level-deep references to sibling skills (e.g., ../aris-research-lit/SKILL.md). Matches 'Good structure; most content is appropriately placed; minor organization gaps' — the ~50-line Phase 6 report template and long prompt blocks could be moved to a references/ file to slim the main file.

Specificity

"Generate and rank research ideas given a broad direction" names the domain and two concrete actions (generate, rank), but omits other core capabilities (novelty validation, pilot experiments, ranked report output), matching the anchor 'Names domain and 1-2 concrete actions, but not comprehensive'. It is not a 4 because it does not list several specific actions.

Completeness

Explicitly answers both: what ("Generate and rank research ideas given a broad direction") and when ("Use when user says '找idea', 'brainstorm ideas', ... or wants to explore a research area for publishable directions") with concrete trigger phrases, matching the top anchor exactly.

Trigger Term Quality

Includes five natural trigger phrases — "找idea", "brainstorm ideas", "generate research ideas", "what can we work on", "explore a research area for publishable directions" — covering colloquial and bilingual variants. Falls short of 5 because common synonyms like "research topics", "paper ideas", or "what should I work on" are missing.

Distinctiveness Conflict Risk

The research-idea niche with distinct triggers (e.g., "generate research ideas", "找idea") is mostly distinct from other skills, but "brainstorm ideas" is generic enough to overlap with a general brainstorming skill — minor overlap risk, matching anchor 4 rather than the minimal-conflict anchor 5.