What do the metrics mean?

Agent call count, Avg duration, and Success rate quantify how often agents are invoked, how long tasks take, and how often they complete successfully.

How do I generate a full analytics report?

Run the analytics with --report full or simply use the default behavior to produce the full report. You can also use --period and --agent for scoped analysis.

What if there is no data yet?

If the work-tracker.jsonl file is missing, you will see the message No work tracker log found. Events will be recorded automatically during Claude Code sessions.

team-shinchan:analytics

npx machina-cli add skill seokan-jeong/team-shinchan/analytics --openclaw

Files (1)

SKILL.md

2.7 KB

Analytics Skill

Analyze work tracker events for observability insights.

Usage

/team-shinchan:analytics                        # Current session summary
/team-shinchan:analytics --report full          # Full analytics report
/team-shinchan:analytics --agent bo             # Single agent detail
/team-shinchan:analytics --trace trace-xxx      # Trace timeline
/team-shinchan:analytics --period 7d            # Last 7 days only

Arguments

Arg	Default	Description
`--report full`	(session)	Generate full analytics report
`--agent {name}`	(all)	Detail for a single agent
`--trace {id}`	(none)	Timeline for a specific trace ID
`--period {N}d`	(all)	Filter events to the last N days
`--format table`	json	Output as text table instead of JSON

Process

Step 1: Locate Tracker File

Use .shinchan-docs/work-tracker.jsonl as the data source.

If file does not exist:

No work tracker log found.
Events will be recorded automatically during Claude Code sessions.

Step 2: Run Analytics Script

Execute the analytics script via Bash:

node "${CLAUDE_PLUGIN_ROOT}/src/analytics.js" .shinchan-docs/work-tracker.jsonl [options]

Map user arguments to script flags:

--report full → no extra flags (full report is default)
--agent <name> → --agent <name>
--trace <id> → --trace <id>
--period <N>d → --period <N>d
Default (no args) → --format table

Step 3: Display Results

For --format table or default: display the text table output directly
For JSON output: format and display key sections with headers

Step 4: Interpret Results

After displaying raw output, provide a brief interpretation:

Highlight busiest agents
Note any agents with low success rates
Summarize delegation patterns
Flag anomalies (e.g. agents with 0 completions)

Metrics Computed

Metric	Description
Agent call count	How many times each agent was invoked
Avg duration	Average time from agent_start to agent_done
Success rate	Ratio of agent_done to agent_start
Session stats	Events, file changes, delegations per session
Event distribution	Percentage breakdown of event types
Delegation graph	Which agent delegated to which, with counts

Important

Analytics script: src/analytics.js (Node.js, built-in modules only)
Data source: .shinchan-docs/work-tracker.jsonl
Trace IDs: generated per user prompt via trace-init hook

Source

git clone https://github.com/seokan-jeong/team-shinchan/blob/main/skills/analytics/SKILL.mdView on GitHub

Overview

The Analytics Skill processes work-tracker JSONL data to extract observability metrics. It highlights agent activity, session stats, event distributions, and delegation patterns to help optimize workflows and detect anomalies.

How This Skill Works

The skill reads the .shinchan-docs/work-tracker.jsonl data source, runs the Node.js analytics.js script, and computes metrics such as agent call count, average duration, and success rate, plus session stats, event distributions, and a delegation graph. Output can be rendered as a JSON report or a human-friendly table, with optional period filtering and agent scope.

When to Use It

Auditing agent performance over a period to identify busy or underperforming agents
Investigating workload distribution and delegation chains across sessions
Detecting anomalies such as agents with 0 completions or unusually long durations
Generating a full analytics report for stakeholders or drills into trace timelines
Inspecting a single agent's details using --agent to isolate behavior

Quick Start

Step 1: Ensure the work tracker data exists at .shinchan-docs/work-tracker.jsonl
Step 2: Run the analytics script, for example: node "${CLAUDE_PLUGIN_ROOT}/src/analytics.js" .shinchan-docs/work-tracker.jsonl [options]
Step 3: View results in table (default) or JSON (--format json) and interpret according to the guidance

Best Practices

Run a full report first to establish baseline metrics
Filter by --period to analyze trends over time (e.g., last 7d)
Review the event distribution to spot rare or high-frequency events
Cross-check the delegation graph with expected workflows
Use JSON output for automation and downstream analytics

Example Use Cases

Monthly digest showing top agents by call count and average duration
Delegation graph illustrating hand-offs between agents
Last 7 days: busiest agents and event type distribution
Identify agents with persistently low success rates
Flag anomalies such as sudden spikes in events or 0 completions

Frequently Asked Questions

Add this skill to your agents