credit-analysis

固收与信用分析：信用债评级、利差分析、违约风险评估、城投债研究、可转债定价与策略。

Quality

63%

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

Fix and improve this skill with Tessl

tessl review fix ./agent/src/skills/credit-analysis/SKILL.md

Quality

Content

46%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 skill's repo-specific value is real and well executed — verified quantlib code paths, unit conventions, and hard-won pitfalls (lookahead-biased percentile, unit mismatches, X4 denominator) — but it is buried in a 900-line monologue whose first half is textbook credit/fixed-income theory Claude already knows. Splitting into reference files and cutting the derivations would roughly halve the token cost at no loss of utility.

Suggestions

Split the monolith into one-level-deep references (e.g. references/china-market.md for §6, references/models.md for §1/§3 derivations, references/quick-ref.md for §8), keeping SKILL.md to the 适用场景 triggers plus §5 quantlib usage — this addresses both progressive_disclosure and conciseness.

Cut or compress the textbook derivations (Merton BS formulas, duration/convexity definitions, yield-curve shape tables) to bare formulas with a pointer to the quantlib function that implements them.

Add explicit validation checkpoints to the analysis sequences (e.g. §6.2 and §5.4): 'before quoting a spread, confirm input_unit against the series' and 'before reporting EDF, cross-check the band via edf_reference_band' to lift workflow_clarity.

Dimension	Reasoning	Score
Conciseness	Sections 1.2–1.4, 3.1–3.3, and 4 extensively re-derive textbook finance Claude already knows (Merton BS equations, Macaulay/modified duration definitions, yield-curve shape tables), making the ~900-line body noticeably padded. Not 1 because §5's quantlib usage and caveats genuinely earn their tokens.	2 / 5
Actionability	§5 and §2.2 give copy-paste-ready imports with real signatures, expected outputs ('bond_price(...) # 104.4518', 'fit.params', 'z.zone') and explicit unit conventions. Not 5 because sections 1–4 and 6 describe frameworks rather than instruct, leaving gaps; not 3 since the executable core is complete and verified against tests.	4 / 5
Workflow Clarity	Sequences exist (§5.2 fitting guidance, §6.2 numbered primary/secondary-market steps, §5.4 unit checks before feeding data), but validation checkpoints are implicit rather than explicit, and there is no end-to-end analysis workflow. Not 4 because no explicit check/verify steps; not 2 because steps are coherent and well-ordered.	3 / 5
Progressive Disclosure	No references/, scripts/, or assets/ exist; the entire ~900-line body is a monolithic SKILL.md where market-structure tables (§6), the formula quick reference (§8), and model derivations clearly belong in separate files. Not 1 because clear headers and §7 cross-links to sibling skills keep it navigable.	2 / 5
	Total	11 / 20 Passed

Description

66%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 names a distinct, well-scoped fixed-income/credit domain with several natural Chinese trigger terms, but it is a pure capability list with no explicit 'when to use this skill' guidance, which limits discoverability and caps completeness. Adding a trigger clause would materially improve it.

Suggestions

Add an explicit 'Use when...' trigger clause, e.g. 'Use when the user asks about 债券定价、久期/凸性、信用利差、违约风险 or 城投债分析' to lift completeness.

Convert some domain nouns into concrete actions (e.g. '计算 YTM/久期/DV01', '估算违约概率与信用利差') to raise specificity and add the missing trigger terms at the same time.

Clarify the boundary with the convertible-bond skill in the description itself (e.g. scope limited to the pure-bond leg), reducing distinctiveness overlap.

Dimension	Reasoning	Score
Specificity	Lists five concrete sub-domains ("信用债评级、利差分析、违约风险评估、城投债研究、可转债定价与策略"), which is several specific capabilities with minor gaps. Not 5 because they are domain nouns rather than concrete actions (no 'compute YTM/duration' style verbs), and not 3 because coverage is broad and domain-specific.	4 / 5
Completeness	The 'what' is clear ("固收与信用分析：信用债评级、利差分析…"), but there is no 'Use when...' or equivalent explicit trigger clause, capping completeness at 3 per the rubric guideline.	3 / 5
Trigger Term Quality	Natural keywords a Chinese fixed-income user would say — 信用债, 利差分析, 违约风险, 城投债, 可转债 — give good coverage. Not 5 because common variations users also say (久期, YTM, ABS, 收益率曲线) are missing; not 3 since the terms present are natural phrases, not jargon-only.	4 / 5
Distinctiveness Conflict Risk	A clear fixed-income/credit niche with distinct triggers, but "可转债定价" overlaps the convertible-bond skill referenced in the body, leaving minor overlap risk. Not 5 for that overlap; not 3 because the core credit-bond triggers are unambiguous.	4 / 5
	Total	15 / 20 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (910 lines); consider splitting into references/ and linking	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	14 / 16 Passed

Repository: HKUDS/Vibe-Trading
Path: agent/src/skills/credit-analysis/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.