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

89%

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

92%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.

An exemplary lean, high-signal skill body: a dense correction table, concrete query modifications, and a two-checkpoint workflow with zero padding. The only soft spot is a few corrections that stay at the direction level rather than giving concrete query patterns (e.g., how to actually search JP/KR/TW sources in-language).

Suggestions

Make the English-bias correction concrete with an actual query pattern or source list (e.g., 'search site:nikkei.com / KIND / TWSE filings with translated terms') the way the leader-bias row does.

Tighten the narrative-bias correction from 'look at the actual product, unit economics, and financial statements' to a checkable action (e.g., 'pull the segment revenue table and label the AI-attributed share before citing it').

Dimension	Reasoning	Score
Conciseness	The ~25-line body is lean and dense: a five-row table where every row adds non-inferable guidance, a four-step application list, and a three-line pairing note. It even assumes Claude's competence ('Force a Munger inversion') without explaining it, and contains no concept explanations or padding.	5 / 5
Actionability	Most guidance is copy-paste ready: '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 as "possibly stale"'. A few corrections remain directional ('look at the actual product, unit economics', 'search JP/KR/TW markets in their own languages' gives no concrete query pattern), which keeps it just below fully executable.	4 / 5
Workflow Clarity	'How to apply' lays out a clear pre-search to post-research sequence with explicit checkpoints on both ends: step 2 (write the thesis, ask 'am I about to fall into this?') and step 4 ('After research, before writing conclusions, re-check: did I cite any disconfirming evidence? did I miss a non-English player? is any key figure stale?'). No destructive or batch operations, so no validation cap applies.	5 / 5
Progressive Disclosure	A single-file skill under 50 lines with well-organized sections (bias table, how-to-apply, layer-pairing map) and no bundle files to reference. Per the rubric's simple-skill guideline this earns a 5; nothing that belongs in a separate file is inlined, and the cross-skill pairings are cleanly layered (data/output/thinking).	5 / 5
	Total	19 / 20 Passed

Description

83%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: it names the domain, lists the concrete checklist contents, gives explicit timing/trigger guidance with natural user phrases, and carves out a distinct reasoning-layer niche. Voice is appropriately third-person/imperative with no padding.

Suggestions

Add a couple of common trigger synonyms (e.g., 'equity research', 'due diligence', 'stock pitch') alongside 'stock screen / sector study / company deep-dive' to broaden natural-match coverage.

Name the sibling skills inline (e.g., 'unlike financial_rigor (numbers) or report_audit (output), this checks the reasoning') to further reduce overlap with a general investment-research skill.

Dimension	Reasoning	Score
Specificity	The description enumerates all five biases with concrete parentheticals ('leader-bias (only big caps)', 'English-bias (miss JP/KR/TW players)') and states exactly how the skill is used ('Load this first, then research with the corrections in mind'). It falls short of 5 only because these are checklist contents rather than multiple concrete executable actions.	4 / 5
Completeness	The 'what' is fully explicit (a self-bias checklist naming all five biases) and the 'when' is explicit trigger guidance equivalent to 'Use when...': 'to run at the START of any investment research task (stock screen / sector study / company deep-dive)', with concrete trigger tasks. Both are answered clearly and specifically.	5 / 5
Trigger Term Quality	Natural phrases a user would actually say are present: 'stock screen', 'sector study', 'company deep-dive', 'investment research'. A few common synonyms ('equity research', 'due diligence', 'stock pitch', 'investment analysis') are missing, keeping it below a 5.	4 / 5
Distinctiveness Conflict Risk	A clear niche (reasoning-layer bias correction for investment research) that is distinct from data- or output-verification skills. 'Investment research' as a trigger still overlaps a general investment-research skill's space and the description does not name its siblings to self-distinguish, so it is not a 5.	4 / 5
	Total	17 / 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: 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.