skills.rest
estebandegetau
estebandegetauCOMMUNITY
Explore all skills by this author →
Website𝕏@estebandegetauGitHub

review-iteration

Diagnose pipeline run failures and map error patterns across stages and codebooks.

1|1|Updated Nov 20, 2025
Search Tags:#pipeline#diagnostics#read-only#error-analysis#targets#codebook
One-click install
npx skills add https://github.com/estebandegetau/Fiscal-shocks --skill review-iteration
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: review-iteration
Source: https://github.com/estebandegetau/Fiscal-shocks/tree/main/.claude/skills/review-iteration
Command: npx skills add https://github.com/estebandegetau/Fiscal-shocks --skill review-iteration

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a structured, read-only analysis of pipeline run results to help users understand failures, map recurring error patterns, and identify next investigative steps without modifying any files or code.

Core Features & Use Cases

  • Structured Diagnosis: Aggregates pipeline outputs, codebook prompts, iteration logs, and success criteria to produce a concise pass/fail summary and root-cause hypotheses.
  • Stage-specific Analysis: Performs tailored analyses for S1 behavioral tests, S2 zero-shot evaluation, and S3 error/ablation studies with example deep-dives and confusion patterns.
  • Investigation Guidance: Ranks ablations and suggests high-leverage follow-ups so users can prioritize experiments or codebook edits.
  • Use Case: After running the targets pipeline, run this Skill to determine which tests failed, why they failed, and what to investigate next before recording changes.

Quick Start

Ask the Skill to analyze codebook c1 at stage s2 and summarize failing metrics, dominant error patterns, and ranked next investigations.

Frequently Asked Questions about review-iteration

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I diagnose R targets pipeline failures without modifying code?▼

Run read-only diagnostics on your R targets pipeline to map error patterns, aggregate pass/fail metrics, and recommend non-destructive next steps without modifying any codebook prompts or project files.

What is the best way to analyze failing metrics in a staged codebook evaluation?▼

Perform stage-specific analysis on S1 behavioral tests, S2 zero-shot evaluations, and S3 ablation studies to surface dominant error patterns and rank high-leverage follow-up investigations.

Can I use read-only diagnostics on targets pipeline outputs accessed via tar_read?▼

Yes, you can use read-only diagnostics on targets pipeline outputs by accessing target data via tar_read or Rscript, alongside codebook prompt files and iteration logs, without altering any project files.

How do I map error patterns across codebooks c1 through c4 in an R pipeline?▼

Apply structured read-only diagnosis to codebook-specific evaluation outputs across stages s1, s2, and s3 to aggregate pass/fail summaries, generate root-cause hypotheses, and map recurring error patterns.

Does pipeline error analysis suggest which codebook edits to investigate next?▼

Yes, pipeline error analysis ranks ablations and suggests high-leverage follow-up investigations so you can prioritize experiments or codebook edits before recording any changes.

Related Skills

Explore More Data & Analytics Skills→
estebandegetauestebandegetau
1

progress-report

Generate a stakeholder-ready Quarto progress report from reports, iteration logs, git commits, and R targets.

Community
Advanced
Ocean326Ocean326

research-training-and-ablation-loop

Analyze AI training runs and ablation results to recommend next experiments.

Community
Advanced
estebandegetauestebandegetau
1

doc-sync

Synchronize project documentation with implementation changes by editing Tier 1 files and logging deltas.

Community
Intermediate
seo-jinseokseo-jinseok

moai-lang-r

Develop R 4.4+ data pipelines, visualizations, and Shiny applications.

Community
Advanced