Foundation Transparency Auditor 2.0

Problem

Foundation audit evidence is often scattered across official registries, websites, and social accounts, with inconsistent disclosure dates and hard-to-verify financial ratios or reputational risk. This skill treats “Is this foundation transparent?” as a structured audit task rather than a subjective label, making the evidence checkable, scoreable, and reviewable.

How it works

• Weighted multi-source collection: It prioritizes China Charity data, then verifies the foundation's official website and WeChat public account, combining scores at 50%, 30%, and 20% weights.
• Five-dimension scoring: It evaluates governance, financial compliance, project execution, disclosure, and net sentiment, producing a 120-point result with a deduction pool for serious adverse events.
• Financial and timeliness checks: It reviews annual reports, audit reports, fundraising costs, administrative expenses, and flags disclosures older than three months.
• Reputational risk scan: It searches for penalties, warnings, negative news, and abnormal operations, grading adverse events into A/B/C levels and applying limited positive offsets.
• HTML report generation: The output is an audit report with score summary, channel details, governance status, financial labels, and an improvement checklist.

Boundaries

It is useful for screening public disclosure, preparing compliance due-diligence drafts, and structuring foundation reviews. It does not replace regulatory conclusions: sentiment findings require cross-verification, unconfirmed claims need cautious wording, and financial ratios still require original statements and applicable law.