P2p Lending Data

验证 Frappe Lending 贷款模块核心流程，包括贷款申请创建、放款计划生成、还款处理及结清退款的自动化测试能力。

This skill is internally inconsistent: its name/description say it's for Frappe Lending tests, but many instructions, preconditions and the human summary talk about ZVT quant backtests, data recorders and trading semantic locks. Before installing or using it, ask the publisher: 1) Which domain is this for (lending tests or ZVT trading/backtest)? 2) Provide a clear, minimal list of required binaries/env vars (Python version, zvt, ZVT_HOME, any recorder credentials) and an explicit install recipe if the skill needs to install packages. 3) Confirm whether the skill will run Python commands/recorders on your host and whether it will write to ~/.zvt or other directories. 4) If you only want lending test automation, request a trimmed SKILL.md that removes ZVT/backtest triggers and semantic locks. Because of the ambiguity, avoid granting broad runtime privileges or running it unattended until the author clarifies these inconsistencies.

Type: OpenClaw Skill Name: p2p-lending-data Version: 0.3.3 The bundle exhibits a critical functional discrepancy: while titled and described as a 'P2P Lending' and 'Frappe' testing tool, the underlying execution logic, pipeline, and 'Doraemon' persona are entirely focused on quantitative trading using the 'zvt' library for A-shares. This mismatch between the stated purpose in SKILL.md and the operational logic in seed.yaml creates a high risk of the AI agent performing unintended financial modeling or trading actions. The bundle enforces strict 'Semantic Locks' (SL-01 to SL-12) and 'Fatal Constraints' (finance-C-*) that steer agent behavior toward trading, which, combined with the functional confusion, makes the bundle deceptive and unreliable.