alpha-zoo

Browse and bench the bundled alpha zoos — prebuilt cross-sectional factor libraries (Kakushadze 101, GTJA 191, Qlib 158, Fama-French / Carhart). Use when the user asks "which alphas exist", wants metadata on a named alpha, or wants to run IC/IR on a whole zoo over a universe.

Quality

87%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

82%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A tight, highly actionable routing skill: the decision tree and pitfalls sections give concrete, executable guidance for every common request, and error conditions are pre-interpreted. The main gaps are dangling reference paths (referenced files absent from the bundle), internal W2/W3/W4 roadmap jargon, and no explicitly sequenced multi-step bench workflow.

Suggestions

Remove or relocate the internal roadmap asides ("W2 scaffold", "zoos pending W3 porting", "lands in W4") — they are developer context, not operational guidance, and trim the token budget without aiding execution.

Add a short sequenced workflow for the bench path (check action=health -> list_alphas to pick targets -> alpha_bench -> interpret the HTML report) so multi-step requests have explicit checkpoints.

Either bundle the referenced files (docs/alpha-zoo/spec.md, src/factors/registry.py, src/factors/factor_analysis_core.py) or remove the dangling Reference section, since none of these paths resolve within the skill directory.

Dimension	Reasoning	Score
Conciseness	The body is lean — tables, a decision tree, and constraints with no explanation of concepts Claude already knows. Minor trimmable content remains: internal roadmap asides ("that's the W2 scaffold — the data pull lands in W4", "zoos pending W3 porting") and a Purpose section that restates the description. Anchor 4 rather than 5 because of these; not 3 since the padding is minor.	4 / 5
Actionability	The decision tree maps concrete requests to copy-paste-ready parameter combinations (e.g. "alpha_bench with zoo=gtja191, universe=csi300, period=2020-2024"), the tools table enumerates the full action space, and pitfalls give exact tokens (equity_cn, mutually exclusive alpha_id/zoo). For an instruction-only skill this fully matches the executable, common-case-covering anchor 5.	5 / 5
Workflow Clarity	Decision-tree routing is unambiguous per request type, and error-recovery guidance exists (action=health surfaces loaded/failed/error reasons; interpretation of empty registry and unimplemented universe loaders). However there is no sequenced multi-step workflow (e.g. health -> list -> bench -> interpret report) and no validation checkpoint around report generation, matching anchor 4 rather than 5.	4 / 5
Progressive Disclosure	A well-signaled, one-level-deep Reference section (docs/alpha-zoo/spec.md, src/factors/registry.py, src/factors/factor_analysis_core.py) with appropriately concise inline content. However none of the referenced files exist in the skill bundle (no references/, scripts/, or assets/ directories), so the pointers are dangling and navigation would fail — matching anchor 4 rather than 5.	4 / 5
	Total	17 / 20 Passed

Description

87%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description: concrete actions, an explicitly named domain, and natural quoted trigger phrases covering both browse and bench use cases. The only weakness is slightly incomplete action coverage (ad-hoc factor analysis and registry health are in the tools but not the description).

Dimension	Reasoning	Score
Specificity	"Browse and bench the bundled alpha zoos — prebuilt cross-sectional factor libraries (Kakushadze 101, GTJA 191, Qlib 158, Fama-French / Carhart)" names the domain and several concrete actions with named libraries. Minor gaps in coverage (no mention of registry health or ad-hoc factor analysis) keep it below the comprehensive anchor 5; it exceeds anchor 3 because it lists more than 1-2 specific actions with a fully named domain.	4 / 5
Completeness	It explicitly answers both what ("Browse and bench the bundled alpha zoos...") and when ("Use when the user asks 'which alphas exist', wants metadata on a named alpha, or wants to run IC/IR on a whole zoo over a universe") with concrete quoted trigger phrases, matching the anchor 5 example pattern exactly.	5 / 5
Trigger Term Quality	Trigger phrases like "which alphas exist", "metadata on a named alpha", and "run IC/IR on a whole zoo over a universe" are natural phrasings a quant user would say, plus named zoo keywords. A few natural variants (e.g. "factor screening", "backtest factors") are missing, matching anchor 4 rather than the comprehensive synonym coverage of anchor 5.	4 / 5
Distinctiveness Conflict Risk	"Prebuilt cross-sectional factor libraries (Kakushadze 101, GTJA 191, Qlib 158, Fama-French / Carhart)" carves out a clear niche with distinct, domain-specific triggers; minimal overlap risk with generic document or code skills.	5 / 5
	Total	18 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: HKUDS/Vibe-Trading
Path: agent/src/skills/alpha-zoo/SKILL.md
Commit: 251b094

Reviewed: 3 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.