CtrlK
BlogDocsLog inGet started
Tessl Logo

thesis-drift

AI Berkshire skill: 投资论文漂移检测:分清事实变化与措辞变化. Source: skills/thesis-drift.md.

59

Quality

67%

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/thesis-drift/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

81%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-sequenced, highly actionable workflow with explicit validation checkpoints, concrete CLI commands, and verbatim output templates. Its main weaknesses are minor redundancy across sections and a monolithic structure that could offload templates/tables into reference files.

DimensionReasoningScore

Conciseness

The body is dense and purposeful with no padding of concepts Claude already knows, but the Improved/Unchanged/Weakened trio and the 5-dimension table recur near-verbatim across A4, A6, the report template, and 关键原则, and the three-mode dispatch is stated in 第一步 then re-expanded — minor trims are available.

4 / 5

Actionability

Provides exact financial_rigor.py subcommands with full flags, fixed input-format specs, a verbatim 8-section output template, and a literal missing-baseline output block; the many unparameterized {…} placeholders across every command are the minor gap keeping it from copy-paste-ready 5.

4 / 5

Workflow Clarity

Clear sequenced flow (mode-dispatch → A1–A6) with explicit validation checkpoints (cross-company stop, 结构缺失/不能编造, no-evidence → Unchanged, B2 mismatch guard, Mode C clean abort), feedback loops, and the fixed dimension table plus output template acting as checklists.

5 / 5

Progressive Disclosure

No bundle files exist and the content is one monolithic 220-line file with no one-level-deep references; headers organize it well and there are no nested reference chains, but the inlined full report template and dimension tables are content that progressive disclosure would split into reference files.

4 / 5

Total

17

/

20

Passed

Description

53%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 states a clear, specific niche and its core distinguishing principle, but omits any 'when to use' trigger guidance and leans on meta-label prefix/metadata rather than enumerated concrete actions. It is functional but below the standard set by the rubric's good examples.

Suggestions

Add an explicit 'Use when...' clause naming natural trigger phrases (e.g. thesis drift, 报告对比, 基本面变化 vs 措辞变化) so Claude knows when to invoke the skill.

Replace the "AI Berkshire skill:" prefix and "Source: skills/thesis-drift.md." metadata with concrete actions the skill performs (e.g. 'Compares two research reports, verifies valuation numbers with financial_rigor.py, and classifies per-dimension drift').

Include common synonyms and both Chinese and English trigger terms (论文漂移/thesis drift, 报告对比) to improve trigger-term coverage.

DimensionReasoningScore

Specificity

Names the domain ("投资论文漂移检测") and one concrete framing principle ("分清事实变化与措辞变化"), but lists no concrete workflow actions, and the "AI Berkshire skill:" prefix is meta-label fluff rather than a capability.

3 / 5

Completeness

Provides a clear 'what' (detect thesis drift, separate fact-change from wording-change) but no 'when' / 'Use when...' trigger guidance, which caps completeness at 3 per the rubric.

3 / 5

Trigger Term Quality

Includes the domain-natural phrase "投资论文漂移" and its gloss "事实变化/措辞变化", but offers no synonyms or variants (e.g. "论文对比", "thesis drift", "报告对比") that a user would naturally say.

3 / 5

Distinctiveness Conflict Risk

"投资论文漂移检测" is a specific niche unlikely to conflict with unrelated skills; minor overlap only with the sibling /thesis-tracker skill it depends on, plus some noise from the "AI Berkshire skill:" prefix and "Source:" metadata.

4 / 5

Total

13

/

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.

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

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.