AgentSkillsCN

phoenix-waterfall-ledger-semantics

针对 Phoenix VC 基金模型，深入解析瀑布式账簿语义与回扣机制。适用于开发层级或账簿式瀑布代码、处理回扣行为，或编写瀑布式真实用例时使用。

SKILL.md
--- frontmatter
name: phoenix-waterfall-ledger-semantics
description:
  'Waterfall ledger semantics and clawback behavior for the Phoenix VC fund
  model. Use when working on tier or ledger waterfall code, clawback behavior,
  or waterfall truth cases.'

Phoenix Waterfall Ledger Semantics

You own GP/LP waterfall semantics for both the tier-based and ledger-based engines, including shortfall-based partial clawback and cross-validation rules.

When to Use

Invoke this skill when:

  • •Editing server/analytics/waterfall-tier.ts or server/analytics/waterfall-ledger.ts
  • •Working on clawback behavior or validating scenario L08 and related cases
  • •Expanding or debugging waterfall truth-case JSONs
  • •Syncing waterfall docs/JSDoc with actual behavior

Files You Own

  • •Implementation:
    • •server/analytics/waterfall-tier.ts
    • •server/analytics/waterfall-ledger.ts
  • •Types / config:
    • •shared/types/waterfall.ts (or equivalent config types)
  • •Truth cases:
    • •docs/waterfall-tier.truth-cases.json
    • •docs/waterfall-ledger.truth-cases.json
  • •Documentation:
    • •Waterfall section in docs/calculations.md
    • •Clawback JSDoc in the waterfall config types

Core Concepts

Tier vs. Ledger

  • •

    Tier engine:

    • •Simple tiers (return of capital, preferred, carry splits)
    • •No ledger-level clawback
  • •

    Ledger engine:

    • •Full ledger of contributions and distributions
    • •Applies shortfall-based partial clawback:
      • •GP carry is limited by fund-level profit above the LP shortfall
      • •If LPs haven't achieved capital + preferred return, GP carry is reduced proportionally
      • •Not a binary all-or-nothing clawback
      • •Not time-based; uses cumulative fund performance at liquidation

Only for simple, no-clawback, no-recycling cases may tier and ledger results be expected to match.

Clawback Semantics

Ensure implementation and docs agree:

  • •Use a shortfall-based formula:
    • •LP shortfall = (required capital + pref) – actual LP distributions
    • •GP clawback = min(GP carry, LP shortfall)
  • •JSDoc must clearly state:
    • •"Shortfall-based partial clawback (NOT hard floor, NOT time-based)"
    • •Include a concrete example referencing the relevant truth case (e.g., L08)

Truth-Case Workflow

  1. •

    Identify failing scenario(s) in tier/ledger truth-cases JSON.

  2. •

    Run targeted tests:

    bash
    npm test -- tests/truth-cases/runner.test.ts -t "<scenario-id>"
  3. •

    Compare:

    • •Expected vs actual totals (LP distributions, GP carry, clawback)
    • •Row-level distributions for ledger scenarios (e.g., quarterly rows)
  4. •

    Decide:

    • •JSON wrong → fix truth case and re-run
    • •Implementation wrong → adjust code, keep semantics aligned with docs and plan

Documentation Sync

Whenever you change behavior:

  • •Update:
    • •docs/calculations.md waterfall section
    • •JSDoc for waterfall config and key functions
  • •Include:
    • •Plain-English description of clawback behavior
    • •Example using a real truth-case ID

Invariants

  • •Clawback must never cause LPs to receive less than capital + pref while GP receives carry.
  • •Ledger engine must respect the LPA semantics described in the plan and JSDoc.
  • •Tier engine is not used to validate clawback scenarios.