Endowment Agent Investment — Security skill for Claude Code
An AI agent given a real mandate, an audit obligation and a point-in-time data wall produces a tactical asset allocation study for a fictional university endowment.
How to install Endowment Agent Investment
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open davidalmeida90/endowment-agent-investment and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Endowment Agent Investment does
An AI agent given a real mandate, an audit obligation and a point-in-time data wall produces a tactical asset allocation study for a fictional university endowment.
Alternatives in Security
- Mcp-scan (Invariant Labs) — MCP security scanner with proxy mode for real-time scanning without infrastructure changes 1.9k ★
- Tech Debt Skill — Claude Code skill that produces a thorough, file-cited tech debt audit of an entire codebase 590 ★
- Gdpr Review — Audit the codebase (or a given path/feature) for GDPR compliance and produce a prioritized findings report 172 ★
README
Ashcroft University Endowment: tactical asset allocation study
[](https://github.com/davidalmeida90/endowment-agent-investment/actions/workflows/verify.yml) [](LICENSE)
Year ended 30 June 2026. Recommendation for the year to 30 June 2027.
Look at it without cloning
**[davidalmeida90.github.io/endowment-agent-investment](https://davidalmeida90.github.io/endowment-agent-investment)** opens every document the agent wrote. The trustee report, the five-year decision record, the interactive dashboard, and the note on how six concurrent desks behaved as a system.
| Watch the run | youtu.be/33fLjqGuJJo, the office building itself, 9 minutes |
| The write-up | davidariasfinance.com/scripts/agentic-investment-committee, with the full brief |
| What not to trust | `AUDIT.md`, read before quoting any number from this |
Everything here was built in this folder. Nothing is imported from elsewhere on the machine it was written on, no API key is used anywhere, and no paywalled source was consulted. Clone the folder, run the commands below, and you get the same numbers.
**Built by an AI agent from a single brief.** `PROMPT.md` is that brief, unedited, and everything except it, `IPS.pdf` and `MANDATE.md` is the agent's work. It ran on my own machine inside my normal working setup, which already had a global instruction file, installed skills, MCP servers and plugins available. That was not a clean room, so I make no claim that the process reproduces from a bare install. The study does. No keys, no private data, no paywalled sources, and the commands below regenerate every number. `HOW_THIS_WAS_BUILT.
Related Skills
Dbt Agent Readiness
Audit a dbt project for what an AI agent will get wrong if you point it at the data today.
QA Strike Team
An adversarial QA and security review skill for apps you build with Claude Code. You point it at a running app
Decision Audit
Analyzes the decision journal to produce a PM Decision Profile. Extracts patterns across optimization bias, in
Audit Leakage
Audit a model, feature set, or pipeline for data leakage — builds the prediction-time availability table and s
Audit Tenancy
Audit multi-tenancy enforcement — RLS policies, withTenantDb usage, query-time tenant filters, and cross-tenan
Vcp
Vibe Coding Protocal - Security-first protocol for AI-generated code, multiple standards with real-time enforc
Related Agents
Fg Strategy Advisor
Finance Guru Senior Portfolio Strategist (Elena Rodriguez-Park). Strategic asset allocation, tactical implemen
Invest Analyst
Read-only investment analytics subagent. Analyses a stock or crypto asset and returns a structured, confidence
Wealth Advisor
Wealth management specialist for high-net-worth advisory, asset allocation, estate planning, and client portfo
