What is the primary purpose of the Session Analysis skill?

It is designed to analyze conversation histories to identify patterns, frustration signals, success markers, and evolving user requirements.

Does this skill work for real-time monitoring?

No, it is intended for post-interaction analysis and signal extraction rather than real-time monitoring of live messages.

What kind of output format does it generate?

The skill produces a detailed JSON report including a scope summary, a list of extracted signals with confidence levels, and detected behavioral patterns.

Can it identify specific tool-usage patterns?

Yes, the skill includes 'Tool Chain' detection which identifies recurring sequences of tool usage, such as moving from Read to Edit to Bash commands.

How does the skill detect user frustration?

It scans for specific indicators such as corrections ('No, I meant...'), reversions, repetition of requests, and explicit negative feedback to pinpoint friction.

Conversation Session Analysis

Name: Conversation Session Analysis
Author: outfitter-dev

byoutfitter-dev

•

Analíticas y Monitorización

Extracts behavioral insights and detects patterns within AI conversation histories to identify success signals and friction points.

The Session Analysis skill provides a sophisticated framework for auditing interactions between users and Claude. By applying a detailed signal taxonomy—categorizing inputs as success, frustration, workflow, or specific requests—it transforms raw conversation logs into actionable behavioral data. The skill tracks the evolution of tasks, identifies recurring tool-chain sequences, and assigns confidence levels to its findings. This is an essential tool for developers and prompt engineers who need to perform post-mortems on complex sessions to refine agent behavior, capture implicit user preferences, and eliminate repetitive workflow bottlenecks.

Características Principales

01Automated detection of behavioral clusters and temporal escalations

02Structured JSON output containing quotes, timestamps, and action items

03Signal taxonomy extraction for success, frustration, and workflow patterns

042 GitHub stars

05Confidence-weighted analysis (High, Medium, Low) for data reliability

06Stage-based execution tracking from input parsing to report synthesis

Casos de Uso

01Extracting implicit user preferences to inform system prompt updates or documentation

02Analyzing tool usage patterns to optimize sequence logic and reduce latency

03Auditing long-running sessions to pinpoint exactly where a workflow broke down

Características Principales

01Automated detection of behavioral clusters and temporal escalations

02Structured JSON output containing quotes, timestamps, and action items

03Signal taxonomy extraction for success, frustration, and workflow patterns

042 GitHub stars

05Confidence-weighted analysis (High, Medium, Low) for data reliability

06Stage-based execution tracking from input parsing to report synthesis

Casos de Uso

01Extracting implicit user preferences to inform system prompt updates or documentation

02Analyzing tool usage patterns to optimize sequence logic and reduce latency

03Auditing long-running sessions to pinpoint exactly where a workflow broke down