Design and audit causal identification strategies for observational studies.
npx skills add https://github.com/flonat/flonat-research --skill causal-design
Please help me install this Agent Skill. Skill: causal-design Source: https://github.com/flonat/flonat-research/tree/main/skills/causal-design Command: npx skills add https://github.com/flonat/flonat-research --skill causal-design
This Skill assists in designing and auditing identification strategies for observational causal inference, helping users choose the right approach and ensuring the credibility of their findings.
Run the causal-design skill with the path to your project or tex file, specifying the mode: causal-design [project-path or tex-file] [--mode design|audit]
High-intent search queries and answers about installing and using this skill.
To design an identification strategy for an observational study, this Skill uses a guided interview process to analyze your research question and data structure, helping you select the most appropriate causal inference approach.
A causal inference audit evaluates the credibility of an existing identification strategy through a 4-phase check, performing diagnostics and robustness tests on your observational study's identification assumptions.
Yes, you can audit an existing causal research methodology directly from a LaTeX file by running the Skill with the file path and specifying the audit mode to perform the 4-phase check.
To check the credibility of your observational study design, you need to provide your project path or tex file along with a clear analysis of your research questions, data structures, and identification assumptions.
Use design mode when initially selecting an identification strategy for a new observational study, and use audit mode to perform diagnostics and robustness checks on an existing causal inference strategy to ensure credibility.
liangdabiaoRefine raw research questions into structured deep-research prompts.
franklee16Estimate treatment effects and run causal inference diagnostics in Python.
20YN04Perform ADF unit-root tests and GARCH modeling on financial time-series data.
the-thinker0Compute causal effects from observational data using do-calculus and DAGs.
