Anthropic's official open source financial intelligence skill library: ready-made methods for investment banking, stock research, private equity, and wealth managementAnthropic 官方开源金融智能体技能库：投行、股票研究、私募、财富管理的现成打法

What really eats up time in financial analysis is repetitive work: the same set of valuation models, fixed-format investment memos, and research notes that need to be realigned every time.

This warehouse breaks down these tasks into structured skills: what the input is, what the output looks like, and how to unify the caliber are all written in the skill definition. The agent follows the execution, and the analyst changes from "doing it all over again" to "reviewing and signing."

Investment banking direction: sorting out process materials, transaction structures and key points of M&A and issuance; stock research: unifying the research note framework and valuation caliber to avoid different people handing over models of different standards.

Private equity direction: structured drafts of due diligence lists and investment memoranda; wealth management: standardized generation of client plans and portfolio descriptions. Each scene is a set of reusable skills, not a one-time prompt word.

Skill definitions are maintained by Anthropic itself and will be updated along with the model capabilities; the format is consistent with Claude's skill specifications, and it can be recognized by copying the directory into the skill directory. There is no need to write prompt words and adjust the format from scratch.

The skill does not contain any customer data, nor does it capture data online - it is only responsible for solidifying the methodology into executable steps, and the data source is still provided by you within the compliance framework.

The download button provides the complete source code package of the main branch (about 0.89 MB). After unzipping, you can see the skill definition directory and documentation for each position.

Use a client that requires supporting skills (such as Claude client) and corresponding subscription or API quota; before implementing it into actual business, it is recommended to run through a small range of samples to confirm that the output caliber is consistent with the company's internal specifications.

This article is compiled from public sources and ships with the matching resource. Product features and pricing are subject to the official page. Resources are for learning and exchange only — please respect the original license.