CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

portfolio-review

AI Berkshire skill: 组合管理:从"研究公司"到"管理组合". Source: skills/portfolio-review.md.

Quality

64%

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/portfolio-review/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

75%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.

The body is a well-structured, highly actionable instruction-only skill: a clear seven-step workflow with concrete tool commands, formulas, thresholds, and output templates, plus built-in validation through the rigor tool and confidence grading. Its main weaknesses are mild — some duplication between the opening philosophy and closing principles sections, and a few undated or data-dependent details.

DimensionReasoningScore

Conciseness

The body is dense with tables, numeric thresholds, and checklists, and assumes financial literacy rather than explaining basics. Minor over-explanation exists: the Codex adapter note is meta-overhead, and the closing 关键原则 section largely restates the opening 设计理念 section. This fits anchor 4 (efficient, minor instances that could be trimmed) better than anchor 5 (every token earns its place) because of that duplication.

4 / 5

Actionability

Concrete guidance throughout: exact commands ("tools/financial_rigor.py verify-valuation", "tools/financial_rigor.py three-scenario"), explicit formulas ("预期年化 ≈ FCF Yield + 预期增速"), numeric thresholds (第一大持仓 <40%, 现金 10-30%), defined table schemas, and a report-structure template. Minor gaps keep it at anchor 4 rather than 5: the Buffett \"$3,820亿现金\" figure carries no date basis, and the 盈亏 column assumes cost data the supported input formats may not include.

4 / 5

Workflow Clarity

Seven explicitly numbered steps with structured sub-steps (4.1–4.4, 5.1–5.3), validation via verify-valuation, A/B/C information-richness grading with a defined degradation path (C级 → 标注低置信度), and forced-explicit conclusions (\"结论必须明确回答\"). This is analysis/reporting, not a destructive or batch operation, so no cap applies; it sits at anchor 4 rather than 5 because there is no recovery loop if market data cannot be retrieved (only confidence downgrading).

4 / 5

Progressive Disclosure

The skill is fully self-contained — no references/, scripts/, or assets/ bundle exists — with a flat, clearly headed hierarchy (设计理念 → 执行流程 steps → 关键原则) and zero nested references, so navigation is trivial. It exceeds the under-50-lines simple-skill case, so anchor 4 (good structure, content appropriately placed, minor organization gaps) is the best fit rather than 5.

4 / 5

Total

16

/

20

Passed

Description

36%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.

The description is essentially a generated file header — a translated title plus a source-file pointer — with no capability statement and no trigger guidance. A user or agent scanning skill descriptions would have little basis to select it over sibling investment skills, and no basis to know when to invoke it.

Suggestions

Replace the title-style description with a capability statement in third person, e.g.: "对投资组合进行审视与优化:解析持仓、单仓位体检、集中度/相关性/机会成本分析、压力测试,并给出调仓建议和组合报告."

Add an explicit trigger clause: "Use when the user asks to review, optimize, or rebalance their portfolio (持仓/仓位/调仓/组合审视), or provides a holdings list such as \"腾讯30%, 美团20%\"."

Drop the internal metadata (\"AI Berkshire skill:\", \"Source: skills/portfolio-review.md\") from the description — it consumes trigger space without helping selection, and move it to the body if it must be retained.

DimensionReasoningScore

Specificity

The description names the domain ("组合管理" — portfolio management) but lists no concrete actions or capabilities; it reads as a document title ("从\"研究公司\"到\"管理组合\". Source: skills/portfolio-review.md") rather than a statement of what the skill does. It fits anchor 2 (domain named, actions minimal/generic) and not anchor 3, which requires at least 1–2 concrete named actions like reviewing holdings or recommending rebalances.

2 / 5

Completeness

The 'what' is only a thematic title — it never states the skill parses holdings, runs concentration/correlation/opportunity-cost/stress analysis, or produces rebalancing recommendations — and there is no 'when to use' clause at all, which caps completeness at 3. Combined with the vague 'what' this fits anchor 2 (vague what, no when) rather than anchor 3, which requires a clear 'what'.

2 / 5

Trigger Term Quality

"组合管理" is a natural phrase a user might say, and "研究公司"/"管理组合" add some signal, but common user phrasings for this task (持仓, 仓位, 调仓, 组合审视/体检) are missing, and "AI Berkshire skill" is internal jargon users would not naturally say. This matches anchor 3 (some relevant keywords, missing variations/synonyms) rather than anchor 4's good coverage.

3 / 5

Distinctiveness Conflict Risk

The "AI Berkshire skill" framing plus the portfolio-management niche is somewhat specific, but the same suite contains sibling investment skills (the body references /industry-research and /investment-checklist), so a request like "帮我看看持仓" could plausibly route to the wrong one. This matches anchor 3 (somewhat specific, could overlap with similar skills) rather than anchor 4's mostly-distinct profile.

3 / 5

Total

10

/

20

Passed

Validation

100%

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

xbtlin/ai-berkshire

Path

codex-skills/portfolio-review/SKILL.md

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.