lessons

Analyze pipeline session artifacts to propose ETHOS.md and skill changes.

1|1|Updated Mar 26, 2026
One-click install
npx skills add https://github.com/Soria-Inc/soria-stack --skill lessons-soria-inc
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: lessons
Source: https://github.com/Soria-Inc/soria-stack/tree/main/lessons
Command: npx skills add https://github.com/Soria-Inc/soria-stack --skill lessons-soria-inc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Runs structured retrospectives over recent data pipeline work to surface patterns of success and failure, generate evidence-backed proposals for ETHOS.md and skill changes, and keep the team's operating principles aligned with real sessions.

Core Features & Use Cases

  • Evidence-driven analysis: Aggregates recent artifacts, mempalace session transcripts, and meeting notes to categorize runs as clean, corrected, failed, or near-miss.
  • Pattern detection & metrics: Identifies recurring error types, gate effectiveness, and verification hotspots to produce concrete metrics and trends.
  • Actionable proposals: Formats proposed ETHOS.md edits, skill updates, CLI gaps, and open questions, and outputs an artifact for human review without auto-committing changes.
  • Use Case: After several pipeline sessions, run this Skill to produce a retro report that highlights repeated ingestion failures, suggests a new gate, and lists commands the CLI needs to support the fix.

Quick Start

Run a lessons retro for the past two weeks and produce a draft ETHOS.md proposal and a retro artifact for review.

Frequently Asked Questions about lessons

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

FAQPage Schema
How do I run a retrospective on recent data pipeline sessions?

Running a pipeline retrospective aggregates recent artifact files and mempalace session transcripts to categorize runs as clean, corrected, failed, or near-miss. It surfaces recurring failures and successful practices from your recent workflows.

What are pipeline retrospectives and how do they identify recurring errors?

Pipeline retrospectives are structured reviews of recent session artifacts that identify recurring error types, gate effectiveness, and verification hotspots. They produce concrete metrics and trends from your ETL workflow data to highlight patterns of success and failure.

Can I automatically apply ETHOS.md changes after a pipeline audit?

You cannot automatically apply ETHOS.md changes after a pipeline audit. The retrospective process formats proposed edits into an artifact for human review, requiring explicit human approval before committing any changes.

Does the retrospective analysis require access to specific artifact directories?

Yes, the retrospective analysis requires read access to the ~/.soria-stack/artifacts directory and ETHOS.md. It also needs the mempalace_search tool schema available via ToolSearch to analyze your pipeline session data.

What's the best way to propose new gates for repeated ETL ingestion failures?

The best way to propose new gates for repeated ETL ingestion failures is to aggregate recent pipeline artifacts and meeting notes to detect patterns. The retrospective outputs an evidence-backed proposal listing needed CLI commands and gate updates for human review.

When should I not use automated artifact analysis for pipeline improvements?

You should not use automated artifact analysis when you lack local artifact files or mempalace session transcripts. The retrospective requires explicit human approval to commit any proposed ETHOS.md edits or skill updates to your pipeline.