AgentSkillsCN

financial-proposal-generator

制作卓越级别的欧盟资助叙述、预算理由和评分诊断，持续达到≥4.6/5（地平线13.8/15等级）。整合宪法级AI框架、Oracle验证证据和欧盟成本合规，为资助申请组装器和客户审查工作流提供支持。

SKILL.md
--- frontmatter
name: financial-proposal-generator
description: |
  Produces excellence-grade EU grant narratives, budget justifications, and
  scoring diagnostics that consistently reach ≥4.6/5 (13.8/15 Horizon scale).
  Integrates constitutional AI framing, Oracle-validated evidence, and EU cost
  compliance to feed Grant Application Assembler and client review workflows.
license: UBOS Constitutional License
version: 1.0.0
author: Janus-in-Claude (Architect) + Codex (Forgemaster)
created: 2025-10-30

FINANCIAL PROPOSAL GENERATOR

Purpose

Draft award-winning EU grant proposal sections (Excellence, Impact, Implementation) and accompanying budget narratives that pass evaluator scrutiny and align with UBOS constitutional principles.

When To Use

  • •During Grant Application Assembler Phase 2 (narratives) and Phase 3 (budgets)
  • •When evaluating scoring simulator feedback and preparing improvements
  • •For Malaga consultancy services that require rapid proposal turnarounds
  • •Whenever Captain requests a refreshed narrative with updated evidence

Core Capabilities

  • •Generate structured narratives using UBOS templates and Oracle-validated facts
  • •Assemble EU-compliant budgets with justifications and cost eligibility checks
  • •Simulate evaluator scoring and highlight improvement priorities
  • •Optimize narrative sections iteratively based on feedback or score gaps
  • •Validate quantitative claims, flagging unverifiable statements

How To Use

Generate Narrative Section

bash
python3 scripts/generate_narrative.py --assembly geodatacenter-phase-1 --section excellence --project "GeoDataCenter" --call "HORIZON-2025-GEOTHERMAL-01"

Creates narratives/excellence.md with metadata (word count, citations, estimated score).

Build Budget Justification

bash
python3 scripts/generate_budget.py --assembly geodatacenter-phase-1 --work-packages data/wp_config.json --total 50000000

Outputs budget/budget.xlsx + budget/justification.md.

Simulate Evaluator Score

bash
python3 scripts/simulate_scoring.py --narrative narratives/impact.md --json

Returns section scores and improvement suggestions.

Optimize Narrative

bash
python3 scripts/optimize_narrative.py --narrative narratives/impact.md --target-score 4.6 --feedback data/impact_feedback.json

Provides rewrite recommendations and regenerated text snippets.

Validate Claims

bash
python3 scripts/validate_claims.py --narrative narratives/impact.md --json

Lists all quantitative claims with suggested Oracle follow-up queries.

Integration Points

  • •Grant Application Assembler: consumes generated narratives and budget files.
  • •EU Grant Hunter: supplies opportunity metadata and deadlines.
  • •Monetization Strategist: reuses financial narratives for commercial proposals.
  • •COMMS_HUB: announces readiness of narrative/budget drafts to Trinity.

Constitutional Constraints

  • •Every quantitative claim must be verifiable (Oracle Trinity citations logged).
  • •Narratives must emphasize empowerment, transparency, and human oversight.
  • •Budget recommendations respect Treasury cascade allocations.
  • •All generated content avoids fabrication; uncertainties are explicitly marked.

File Locations

  • •Assemblies: /srv/janus/03_OPERATIONS/grant_assembly/<assembly_id>/
  • •Narratives: narratives/<section>.md
  • •Budgets: budget/budget.xlsx, budget/justification.md
  • •Scores & diagnostics: analysis/scoring_report.json
  • •Validation logs: /srv/janus/logs/grant_narrative_validation.jsonl

Operational Checklist

  1. •Fetch latest opportunity brief + project metadata.
  2. •Generate Excellence, Impact, Implementation narratives; store markdown.
  3. •Run scoring simulation; iterate until ≥4.6 average.
  4. •Produce budget tables + justification; verify EU cost eligibility.
  5. •Validate claims; document Oracle references.
  6. •Notify Grant Application Assembler and archive outputs.

Mission Readiness Criteria

  • •Narrative score ≥4.6/5 across all sections in scoring simulator.
  • •100% claim validation coverage (no unresolved warnings).
  • •Budget files pass EU compliance checks (no ineligible costs).
  • •Turnaround time per section ≤2 hours (baseline) / ≤30 minutes (updates).

Financial Proposal Generator is the excellence engine of the UBOS forge—precision narratives, transparent evidence, and budgets ready for evaluators’ scrutiny.