Conciseness

Lean and efficient throughout: the body is a dense endpoint decision table, executable curl snippets, and terse operational rules; it never explains concepts Claude already knows (no 'what is an API key' padding) and every section carries load (e.g. error semantics like '503 not_available... retry later, don't treat as empty data'). The two-line intro only conveys non-obvious guidance about which datasets are distinctive. This matches the level-5 anchor; level 4 would require identifiable over-explanation to trim, and none is present.

Actionability

Fully executable, copy-paste-ready commands: 'curl -s -H "X-API-KEY: $FINTEL_API_KEY" "https://api.fintel.io/v1/securities/us/AAPL/short-volume" | python3 -m json.tool', plus concrete identifier-resolution curls and an explicit `claude mcp add` command. The Step 3 table maps ~20 user intents to exact endpoint paths with parameters and error notes, covering the common cases. Matches the level-5 anchor; no pseudocode or missing execution details.

Workflow Clarity

A clear numbered sequence (resolve key → resolve security → match endpoint → call → MCP alternative → respond) with explicit validation and recovery checkpoints for every failure mode: 'KEY_NOT_SET — ask the user for their key', '403 means not entitled', '503 not_available means retry later; don't treat as empty data', and 'Surface meta.warnings to the user when present'. This is a read-only lookup skill, so no destructive/batch cap applies, and the error-recovery loops meet the level-5 anchor; level 4 would leave some checkpoint implicit, and none is.

Progressive Disclosure

The body is a clear overview that pushes full parameter details to a single one-level-deep reference — 'Full parameter details, country/exchange discovery endpoints, and more curl examples: read `references/api-reference.md`' — and a 'Reference Files' section that says exactly what the file contains. The bundle matches: references/api-reference.md exists and holds the bulk endpoint tables. Clean one-level structure, well-signaled navigation, matching the level-5 anchor.

Specificity

The description enumerates many specific concrete capabilities — 'short interest, short interest ratio or days to cover, short squeeze data, borrow rate or cost to borrow, shares available to borrow, daily short volume, fails-to-deliver (FTD), 13F institutional owners, insider buying or selling from SEC Form 3/4/5' — comprehensive coverage of what the skill queries, matching the level-5 anchor. It is not the level-4 anchor because there are no meaningful coverage gaps; the enumerated datasets span the full surface described in the body.

Completeness

Explicitly answers both questions: what — 'Query Fintel (fintel.io) institutional market data through its REST API or official MCP server' — and when — 'Use this skill whenever the user mentions Fintel or asks about [concrete trigger phrases]'. It also adds a scope boundary ('Read-only: GET endpoints only'), matching the level-5 anchor with concrete trigger phrases; level 4 would require a less explicit 'when' clause.

Trigger Term Quality

Comprehensive natural trigger phrasing including synonyms and near-user-verbatim terms: 'who owns X', 'cost to borrow', 'shares available to borrow', 'short squeeze data', 'fails-to-deliver (FTD)', 'EOD or last prices', and identifier names 'ticker, CUSIP, ISIN, or FIGI'. This matches the level-5 anchor (comprehensive coverage of natural terms including synonyms); nothing common is missing.

Distinctiveness Conflict Risk

Mostly distinct: triggers are anchored on 'Fintel' and niche datasets (borrow rates, FTD, short volume, 13F, Form 3/4/5) that other skills rarely claim. However, several listed triggers — 'analyst price targets, ratings and forecasts', 'dividend or earnings history and calendars', 'EOD or last prices' — are generic market-data terms that would also fire sibling stock-data skills, which is exactly the level-4 anchor ('minor overlap risk with closely related skills'). Not level 5 because the overlap on these generic terms is more than minimal; not level 3 because the provider anchor and majority-niche triggers keep it clearly distinguishable.