domains / families / software-engineering / error-analysis

error-analysis

separate

Help the user systematically identify and categorize failure modes in an LLM pipeline by reading traces. Use when starting a new eval project, after significant pipeline changes (new features, model switches, prompt rewrites), when production metrics drop, or after incidents.

When to use

• starting a new eval project
• after significant pipeline changes
• when production metrics drop
• after incidents
• reading LLM pipeline traces
• building a catalog of system failures
• grouping failures into categories
• computing failure rates

Skill page

Source