CtrlK
BlogDocsLog inGet started
Tessl Logo

private-company-research

AI Berkshire skill: 未上市公司研究:多Agent并行深度研究框架. Source: skills/private-company-research.md.

56

Quality

63%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./codex-skills/private-company-research/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

77%

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

A highly actionable, well-sequenced multi-agent research workflow with strong validation checkpoints, undermined mainly by monolithic length and the absence of any progressive file structure. The main weakness is token efficiency, not clarity or actionability.

Suggestions

Move the six large task-description blocks (任务1–任务6) into separate reference files (e.g., references/task-business.md, task-financials.md) and point to them from a concise overview to cut inline length dramatically.

Tighten or relocate the 'AI研究偏见自觉' conceptual framing so the body leads with the executable workflow and keeps the philosophy one level deep.

DimensionReasoningScore

Conciseness

At ~1080 lines the body is very long and includes conceptual passages (e.g., the 'AI研究偏见自觉' section and unlisted-vs-listed characteristics) plus redundant tables that could be tightened, so it is 'mostly efficient but could be tightened'; not a 1 because most content is specialized actionable methodology rather than concepts Claude already knows, not a 3 because of clear verbosity.

2 / 3

Actionability

Provides concrete task subjects/activeForms, fully-specified task description templates with tables, named search sources, explicit commands (e.g., 'python3 tools/financial_rigor.py', 'run the date command'), and a precise confidence-labeling convention — fully executable, copy-paste-ready guidance matching the top anchor.

3 / 3

Workflow Clarity

A clear nine-step sequence (第一步–第九步) with explicit validation checkpoints in 第六步 (data-conflict arbitration, signal-consistency checks, anti-bias review) and an enforced parallel-launch requirement for the 6 batch agents, matching 'Clear sequence with explicit validation steps; feedback loops for error recovery'.

3 / 3

Progressive Disclosure

Sections are well-organized, but everything is inline in one 1080-line SKILL.md with no bundle files and no file-splitting of the six large task blocks that should be separate references, fitting 'content that should be separate is inline'; not a 1 because organization is good and references are not deeply nested, not a 3 because nothing is offloaded to one-level-deep reference files.

2 / 3

Total

10

/

12

Passed

Description

50%

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

The description names a clear domain and method but omits any usage trigger guidance and leans on a couple of meta fragments ('AI Berkshire skill', 'Source: ...') that are not user-facing triggers. It answers 'what' but not 'when'.

Suggestions

Add an explicit 'Use when...' clause naming natural triggers (e.g., 'Use when the user asks to research, value, or analyze an unlisted/private company such as Ant Group, Xiaohongshu, SpaceX, or Stripe').

Replace the 'AI Berkshire skill:' / 'Source: skills/...md' meta prefix with concrete capabilities (e.g., 'dissects business model, pieces together financials, derives valuation, and maps risks for unlisted companies') to raise specificity and trigger coverage.

Include common trigger variations users would actually say ('private company research', '未上市公司估值', 'pre-IPO valuation', 'unlisted company analysis').

DimensionReasoningScore

Specificity

Names a concrete domain ('未上市公司研究') and a method ('多Agent并行深度研究框架') but does not list multiple specific actions — closest to 'Names domain and some actions, but not comprehensive'; not a 3 because no enumerated concrete capabilities, not a 1 because the domain is named rather than vague.

2 / 3

Completeness

It states what the skill does (private-company deep research framework) but has no 'Use when...' clause or equivalent trigger guidance, so per the judging guideline completeness is capped at 2 ('Has what, but when is missing or only implied').

2 / 3

Trigger Term Quality

'未上市公司研究' is a natural term a user might say, but '多Agent并行' is technical jargon and 'AI Berkshire skill' / 'Source: ...' are metadata, not triggers; only one natural keyword is present, so it stops at 'Some relevant keywords but missing common variations'.

2 / 3

Distinctiveness Conflict Risk

'未上市公司研究' is a fairly distinctive niche, but '多Agent并行深度研究框架' overlaps with generic multi-agent research skills and the description lacks explicit triggers to disambiguate, fitting 'Somewhat specific but could still overlap with similar skills'.

2 / 3

Total

8

/

12

Passed

Validation

93%

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

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

skill_md_line_count

SKILL.md is long (1087 lines); consider splitting into references/ and linking

Warning

Total

15

/

16

Passed

Repository
xbtlin/ai-berkshire
Reviewed

Table of Contents

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.