How are the results formatted?

The output is a detailed JSON object that includes the analysis scope, a categorized list of signals with confidence levels, and a summary of detected patterns and action items.

What specific signals does session-analysis track?

The skill tracks four primary categories: Success (praise, adoption), Frustration (corrections, reversions), Workflow (sequences, transitions), and Request (prohibitions, requirements).

Does this skill work for real-time monitoring?

No, this skill is designed for batch analysis of existing message ranges rather than real-time monitoring or generating new content.

When is the best time to use this skill?

It is best used when you need to analyze a conversation history, identify why a task failed, or extract long-term user preferences from previous iterations.

Can it analyze very short conversations?

Yes, however, for scopes with fewer than five messages, the skill automatically skips the pattern detection phase to jump straight to synthesis.

Session Analysis & Pattern Detection

Name: Session Analysis & Pattern Detection
Author: outfitter-dev

byoutfitter-dev

•

Analytics & Monitoring

Analyzes conversation history to identify behavioral patterns, user preferences, and workflow signals for improved agent performance.

About

The session-analysis skill transforms raw conversation logs into actionable insights by extracting success and frustration signals, mapping workflow transitions, and detecting recurring behavioral patterns. It helps developers understand exactly where a project is succeeding or stalling by providing a structured JSON report that highlights user requirements, tool usage sequences, and confidence-rated signal clusters. By automating the 'what went wrong' process, it enables more precise refinements to agent instructions and project guidelines.

Key Features

Multi-stage signal extraction across Success, Frustration, Workflow, and Request categories
Advanced pattern detection for repetition, evolution, and tool chain sequences
Confidence-rated signal taxonomy to distinguish between explicit and implicit feedback
18 GitHub stars
Temporal tracking to identify project escalation, de-escalation, or cyclical issues
Structured JSON output containing signal logs, pattern summaries, and action items

Use Cases

Optimizing development workflows by identifying recurring sequences of tool usage
Extracting persistent user requirements and coding preferences to build better project rules
Conducting post-mortems on complex agent interactions to identify friction points

About

Key Features

Multi-stage signal extraction across Success, Frustration, Workflow, and Request categories
Advanced pattern detection for repetition, evolution, and tool chain sequences
Confidence-rated signal taxonomy to distinguish between explicit and implicit feedback
18 GitHub stars
Temporal tracking to identify project escalation, de-escalation, or cyclical issues
Structured JSON output containing signal logs, pattern summaries, and action items

Use Cases

Optimizing development workflows by identifying recurring sequences of tool usage
Extracting persistent user requirements and coding preferences to build better project rules
Conducting post-mortems on complex agent interactions to identify friction points