abs-cashflow-modeling

建模资产支持证券交易结构，模拟抵押贷款池现金流、债券分级偿还和瀑布分配，分析 tranche 收益与风险表现。 触发场景：(1) 用户要构建 ABS 交易模型，分析优先级债券和股权层的本息回收；(2) 用户要模拟 LIBOR 浮动利率抵押贷款池，查看利率重设期的现金流变化；(3) 用户要建模含阶梯利率的债券结构，...

Security Scan

The skill mixes ABS cash‑flow modeling with a ZVT trading/backtest pipeline, includes runtime requirements and preconditions that are not declared (eg. ZVT, ZVT_HOME, Python version), and its install script performs global pip installs without isolation — these inconsistencies merit caution before use.

Purpose & Capability

The skill name/description claim ABS cashflow and tranche modeling, but the SKILL.md and human_summary emphasize a ZVT-based A-share trading/backtest pipeline and many trading use cases. The metadata line "Requires Python 3.12+ with uv package manager" and heavy ZVT references are not reflected in the simple install script. The mix of ABS deal modeling and ZVT trading/backtest functionality is plausible for a broad finance package, but the package does not clearly justify or declare ZVT as a required dependency.

Instruction Scope

SKILL.md and seed.yaml instruct the agent to read many local reference files (seed.yaml, references/*) and to run preconditions that execute python -c commands (checking zvt, ZVT_HOME, recorders, etc.). Those preconditions access environment state and can prompt the agent to run pip installs or initialize directories. The skill's declared requirements list no env vars, yet instructions reference ZVT_HOME and other filesystem locations and tell the agent to run commands that modify local data directories — this expands the runtime scope beyond what the manifest declares.

Install Mechanism

The only install mechanism is scripts/install.sh which runs several python3 -m pip install calls for common packages (pandas, numpy, requests, toolz, lenses, graphviz, schema, htpy, dateparser, more-itertools). Packages are from PyPI (no suspicious URLs), but the script installs globally (no virtualenv) and does not install or verify zvt or the declared Python 3.12+ requirement. Global pip installs and lack of verification are a maintenance/security concern.

Credentials

The skill declares no required environment variables, but SKILL.md/seed.yaml/preconditions reference ZVT_HOME and expect zvt to be present. The preconditions include instructions that, on failure, tell the user/agent to run pip install zvt and initialize ~/.zvt. Access to ZVT_HOME and writing to user home are not declared up front — this is disproportionate and inconsistent with the manifest.

Persistence & Privilege

always:false and no explicit modification of other skills. However seed.yaml's execution_protocol mandates re-reading seed.yaml and running preconditions on execute, and preconditions may create/modify user directories (eg. ~/.zvt) or install additional packages. That behavior grants ongoing file-system effects when the agent follows the skill's runtime instructions; it's not automatically privileged but it does enable persistent local state changes.

What to consider before installing

This skill is internally inconsistent in ways worth checking before installing. Specifically: - Ask the owner/maintainer to clarify the primary purpose: ABS cashflow modeling or ZVT trading/backtest, and to reconcile the README/metadata. - The package references ZVT and expects Python 3.12+, but the install script does not install zvt or verify Python version. Expect the agent to prompt you to run python commands (including pip install zvt) and to create a ~/.zvt directory unless you change behavior. - The install script runs pip install globally. Install in an isolated virtual environment (venv/conda) or container to avoid modifying your system Python environment. - The skill will read its seed.yaml and many references at runtime; review those files (seed.yaml, references/*) locally for any commands you don't want an agent to execute. - If you need to install: (1) inspect scripts/install.sh and run it in a controlled venv; (2) confirm whether zvt is required and request the maintainer add it to the install script or document why it's optional; (3) ensure you are comfortable with the skill creating/modifying ~/.zvt or set ZVT_HOME to a safe path. - If you cannot verify the origin (source is unknown and homepage none), prefer running in sandboxed environment or decline until the maintainer provides clearer provenance and an updated install procedure.

I help you build quant strategies on A-share with ZVT — from data fetch to backtest, one flow. Just tell me what you want; I'll write the code, you don't have to dig docs. (Heads up: ZVT natively supports A-share, HK, and crypto. US stocks — stockus_nasdaq_AAPL — are half-baked; don't bother for serious work.)

Pipeline

data_collection -> data_storage -> factor_computation -> target_selection -> trading_execution -> visualization

Top Use Cases (40 total)

Basic ABS Deal Model (UC-001)

Model a basic asset-backed securities deal with mortgage pool, bonds, fees, and waterfall to analyze cashflows and tranche performance Triggers: basic deal, ABS, mortgage pool

Adjustable Rate Mortgage Pool (UC-002)

Model an adjustable rate mortgage pool with LIBOR-based floating rates and periodic resets Triggers: ARM, adjustable rate, LIBOR

Bond Step-Up Rate (UC-003)

Model bonds with scheduled rate step-ups at specific dates for ABS deal structuring Triggers: step-up, bond rate, scheduled increase

For all 40 use cases, see references/USE_CASES.md.

Install

# One-time setup before first use
bash scripts/install.sh

Execute trigger: When user intent matches intent_router.uc_entries[].positive_terms AND user uses action verb (run/execute/跑/执行/backtest/fetch/collect)

What I'll Ask You

• Target market: A-share (default), HK, or crypto? (US stocks in ZVT are half-baked — stockus_nasdaq_AAPL exists but coverage is thin)
• Data source / provider: eastmoney (free, no account), joinquant (account+paid), baostock (free, good history), akshare, or qmt (broker)?
• Strategy type: MACD golden-cross, MA crossover, volume breakout, fundamental screen, or custom factor?
• Time range: start_timestamp and end_timestamp for backtest period
• Target entity IDs: specific stocks (stock_sh_600000) or index components (SZ1000)?

Semantic Locks (Fatal)

ID	Rule	On Violation
SL-01	Execute sell orders before buy orders in every trading cycle	halt
SL-02	Trading signals MUST use next-bar execution (no look-ahead)	halt
SL-03	Entity IDs MUST follow format entity_type_exchange_code	halt
SL-04	DataFrame index MUST be MultiIndex (entity_id, timestamp)	halt
SL-05	TradingSignal MUST have EXACTLY ONE of: position_pct, order_money, order_amount	halt
SL-06	filter_result column semantics: True=BUY, False=SELL, None/NaN=NO ACTION	halt
SL-07	Transformer MUST run BEFORE Accumulator in factor pipeline	halt
SL-08	MACD parameters locked: fast=12, slow=26, signal=9	halt

Full lock definitions: references/LOCKS.md

Top Anti-Patterns (15 total)

• AP-INSURANCE-001: Implicit numeric format assumptions without validation
• AP-INSURANCE-002: Triangle axis construction with invalid temporal ordering
• AP-INSURANCE-003: Cumulative/incremental triangle representation misuse

All 15 anti-patterns: references/ANTI_PATTERNS.md

Evidence Quality Notice

[QUALITY NOTICE] This crystal was compiled from blueprint finance-bp-076. Evidence verify ratio = 37.8% and audit fail total = 22. Generated results may have uncaptured requirement gaps. Verify critical decisions against source files (LATEST.yaml / LATEST.jsonl).

Reference Files

File	Contents	When to Load
references/seed.yaml	V6+ 全量权威 (source-of-truth)	有行为/决策争议时必读
references/ANTI_PATTERNS.md	15 条跨项目反模式	开始实现前
references/WISDOM.md	跨项目精华借鉴	架构决策时
references/CONSTRAINTS.md	domain + fatal 约束	规则冲突时
references/USE_CASES.md	全量 KUC-* 业务场景	需要完整示例时
references/LOCKS.md	SL-* + preconditions + hints	生成回测/交易代码前
references/COMPONENTS.md	AST 组件地图（按 module 拆分）	查 API 时

---

Compiled by Doramagic crystal-compilation-v6.1 from finance-bp-076 blueprint at 2026-04-22T13:00:28.210602+00:00. See human_summary.md for non-technical overview.