earnings-review

AI Berkshire skill: 财报精读：一手资料深度解读. Source: skills/earnings-review.md.

Quality

63%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./codex-skills/earnings-review/SKILL.md

Quality

Content

77%

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 strong, actionable workflow with executable validation commands and an explicit exit-gate feedback loop, scoring high on actionability and workflow clarity. Its main weaknesses are token overhead from motivational/philosophical and meta content, and a monolithic structure with no progressive disclosure via bundle files.

Suggestions

Trim the 设计理念 rationale, the investor quotes, and non-essential Codex-adapter prose to tighten conciseness without losing the operational guidance.

Move the detailed data-source fallback rules (and optionally the MD&A signal taxonomy / footnote checklist) into a reference file under references/ and link to it one level deep, converting the monolithic SKILL.md into an overview.

Specify the JSON schema expected for 'python3 tools/report_audit.py verdict --results' so the exit-gate step is fully copy-paste ready.

Dimension	Reasoning	Score
Conciseness	The body is mostly operational (tables, thresholds, commands) but carries unnecessary tokens — motivational quotes from 李录/巴菲特/段永平, a 设计理念 rationale section, and a Codex adapter meta-note — that could be tightened without losing guidance, matching the 'mostly efficient but includes some unnecessary explanation' anchor.	2 / 3
Actionability	It provides multiple fully executable commands with real example arguments (e.g. 'python3 tools/financial_rigor.py cross-validate --metric revenue --values 108.3e9 107.9e9 …'), concrete numeric thresholds (>100%为佳, <80%需警惕), and specific signal taxonomies, matching the 'fully executable code/commands; copy-paste ready' anchor.	3 / 3
Workflow Clarity	A clear 8-step sequence is fronted by an A/B/C data-availability gate, with explicit validation (financial_rigor.py cross-checks) and an exit-gate feedback loop (准出/打回 → 修正后重审) plus checklists for footnotes and anomaly detection, matching the 'clear sequence with explicit validation steps and feedback loops' anchor.	3 / 3
Progressive Disclosure	The skill is well-sectioned but monolithic at ~230 lines with no bundle files (references/, scripts/, assets/ absent); content that could be split (data-source fallback rules, MD&A signal taxonomy, footnote checklist) is inline, and references like skills/financial-data.md are not well-signaled one-level-deep bundle refs, fitting the score-2 anchor.	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 identifies a clear domain and approach but is thin: it functions more as a titled label plus provenance metadata than a capability statement, and it lacks any explicit 'Use when' trigger guidance. Most dimensions land at the middle anchor as a result.

Suggestions

Drop the non-capability clutter ('AI Berkshire skill:' and 'Source: skills/earnings-review.md.') and lead with a concrete action verb phrase in third person.

Add an explicit 'Use when…' trigger clause naming natural user phrasings, e.g. 'Use when the user asks to analyze, interpret, or close-read an earnings report, 10-K/10-Q, 年报, 季报, or earnings call transcript.'

List 2-3 specific concrete actions (e.g. extract and cross-validate core financials, parse MD&A tone and management commitments, mine footnotes for hidden risks) to raise specificity and distinctiveness.

Dimension	Reasoning	Score
Specificity	The description names a concrete domain (财报) and high-level action verbs (精读, 深度解读), but does not list multiple specific concrete actions, matching the 'names domain and some actions, but not comprehensive' anchor rather than the score-3 list of discrete actions.	2 / 3
Completeness	It conveys what the skill does (财报精读：一手资料深度解读) but provides no 'Use when…' clause or equivalent explicit trigger guidance, so completeness is capped at 2 per the judging guidelines.	2 / 3
Trigger Term Quality	It includes natural terms a user might say (财报, 精读, 一手资料), but misses common variations like earnings, 年报, 季报, or 电话会, and the 'AI Berkshire skill:' and 'Source: skills/earnings-review.md.' fragments add no trigger value.	2 / 3
Distinctiveness Conflict Risk	Financial-report close-reading is a reasonably distinct niche, but the noisy 'AI Berkshire skill:' prefix and provenance metadata dilute the signal and it could still overlap with other financial-analysis skills, fitting the score-2 anchor.	2 / 3
	Total	8 / 12 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: xbtlin/ai-berkshire
Path: codex-skills/earnings-review/SKILL.md
Commit: 634a399

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.