Finance
Journal entries, reconciliations, financial statements, variance analysis and SOX testing for accountants.
Write SQL for your warehouse dialect, profile a new dataset, build charts and dashboards, and check an analysis before you share it.
Free plan: up to 10 private skills and 3 members, no credit card.
A skill is a written procedure your assistant reads and follows. A command is a named shortcut that starts one. These are the names in this plugin.
Examples the plugin names, not a tested list. It reads from the ones you connect and otherwise works on what you paste in. An underlined name has a page in our connector directory.
Counted from the plugin’s own manifest on October 3, 2026.
The free plan includes up to 10 private skills and 3 members, with no credit card.
Choose the plugins your team needs. You can turn one off again at any time.
Your connector URL appears after you sign in. Add it once in your assistant’s connector settings, then ask for the job in plain words.
It writes the query and the analysis, and an analyst checks the result before it is shared. Running queries against live data needs your data warehouse connected. Without that it writes SQL for you to run and works from the files you upload.
Ported from Anthropic's open-source plugins by Agentman. Anthropic does not review, endorse or support this port. Questions about it come to us.
Read what a skill does before you rely on its output, and connect only the tools you mean to share with it.
7 skills, 6 commands, 17 connectors, listed by name on this page. In plain terms: SQL written for your dialect, including Snowflake, BigQuery, Databricks and PostgreSQL; Dataset profiling: shape, quality, nulls, outliers and column distributions; Statistical analysis: descriptive stats, trends, outliers and hypothesis tests; Charts in Python with matplotlib, seaborn or plotly; Self-contained interactive HTML dashboards with charts, filters and tables; A QA pass on methodology, accuracy and bias before an analysis goes out.
It writes the query and the analysis, and an analyst checks the result before it is shared. Running queries against live data needs your data warehouse connected. Without that it writes SQL for you to run and works from the files you upload.
It is ported from Anthropic's open-source plugins by Agentman, so it can be served over an MCP connection. Anthropic does not review, endorse or support the port.
Create an Agentman account, turn the plugin on in your workspace, and add your connector URL to your assistant. The URL appears after you sign in. You set the connection up once, not once per plugin.
Not to start. Its README names 17 tools it can read from when they are connected, such as Snowflake, Databricks and BigQuery. Without them it works on what you paste in or upload.
Create an account, turn the plugin on in your workspace, and add one connector URL to your assistant. Your connector URL appears after you sign in.