research-discipline

A short self-bias checklist to run at the START of any investment research task (stock screen / sector study / company deep-dive). Four biases that systematically warp AI research — leader-bias (only big caps), English-bias (miss JP/KR/TW players), narrative-bias (chase concept labels), confirmation-bias (only bullish evidence) — plus recency-bias. Load this first, then research with the corrections in mind. Not a workflow, just a 60-second attitude reset that materially improves coverage and honesty.

Quality

83%

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

78%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, well-organized instruction-only skill: the bias table delivers specific corrections and the application sequence ends in an explicit self-check. Slight redundancy between the table and the how-to list, plus a couple of implied mechanics, keep it just off maximum on conciseness and actionability.

Suggestions

Remove the mild puffery in the opening ('60 seconds here materially improves coverage and intellectual honesty') or fold it into the description to avoid re-stating it in the body.

Add one concrete pointer for the English-bias correction, e.g. a named source or query-language hint for JP/KR/TW markets, so the instruction is fully executable rather than implied.

Tighten 'How to apply' to reference the table rows instead of rephrasing each bias, reducing token overlap with the table.

Dimension	Reasoning	Score
Conciseness	The body is lean — a compact table plus a 4-step application list and a short 'pairs with' note, assuming Claude's competence with no padding — but the opening sentence restates the description and the 'How to apply' section lightly rephrases the table, so a few tokens could be trimmed, landing just below the 5 anchor.	4 / 5
Actionability	The corrections column gives concrete, executable guidance with literal query additions ('add `small cap` / `mid cap` / `supply chain` to queries', 'search the bear case ("X risks / problems / bear case")', 'Prefer the last 30 days; mark anything older than a year'), covering common cases, but practical mechanics for some steps (e.g. how to actually query JP/KR/TW markets in their own languages) are left implied, a minor gap.	4 / 5
Workflow Clarity	The 'How to apply' section gives a clearly sequenced before-search → thesis → broaden-plan → after-research re-check flow, with an explicit validation checkpoint at step 4 ('did I cite any disconfirming evidence? did I miss a non-English player? is any key figure stale?'); it is not a pure single-action skill and carries only one checkpoint, so it sits just under the 5 anchor.	4 / 5
Progressive Disclosure	At well under 50 lines with no need for external references, the content is organized into clean sections (intro, bias table, application steps, sibling-skill mapping) with no nesting, satisfying the simple-skill exception for a top score.	5 / 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 that concretely names five biases with their symptoms, gives explicit start-of-task trigger guidance, and stays in third person with no fluff. The only minor weakness is that the literal actions (run/load) are generic while the bias content is rich.

Dimension	Reasoning	Score
Specificity	Names the domain (investment research) and enumerates five concrete named biases each with a parenthetical behavior ('only big caps', 'miss JP/KR/TW players', 'chase concept labels', 'only bullish evidence'), but the actual actions ('run at the START', 'Load this first') are slightly generic, leaving minor gaps versus the comprehensive 5 anchor.	4 / 5
Completeness	Clearly answers both what ('A short self-bias checklist... Four biases... plus recency-bias') and when ('to run at the START of any investment research task (stock screen / sector study / company deep-dive)') with concrete trigger phrases, matching the explicit-what-and-when anchor; the equivalent 'Use when...' trigger guidance is present so the 3-cap does not apply.	5 / 5
Trigger Term Quality	Includes natural phrases a research user would actually say — 'investment research task', 'stock screen', 'sector study', 'company deep-dive' — with several synonyms, but a few common variations (e.g. 'equity research', 'pitch') are absent, stopping just short of the comprehensive 5 anchor.	4 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (AI investment-research bias self-check) with distinct triggers (stock screen / sector study / company deep-dive) and explicit differentiation from sibling skills ('Not a workflow, just a 60-second attitude reset'), giving minimal conflict risk.	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/research-discipline/SKILL.md
Commit: 1907e47

Reviewed: 1 day 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.