Generate Analytics Reports

This skill enables terminal-based analytics report generation using the Olakai CLI, eliminating the need to access the web UI for analytics insights.

For full documentation, see: https://app.olakai.ai/llms.txt

Prerequisites

Before generating reports, ensure:

1. CLI is authenticated

olakai whoami

2. You have the agent ID (if reporting on specific agent)

olakai agents list --json | jq '.[] | {id, name}'

Report Generation Workflow

• Gather context - Determine agent ID, date range, and report type

• Query data - Use CLI commands with --json flag

• Process output - Extract relevant metrics using jq

• Generate visualizations - Create ASCII charts and markdown tables

• Present report - Format and display the complete report

Available Data Sources

Command Data Retrieved

olakai activity list --json

Events with tokens, model, risk, status

olakai activity list --include-analytics --json

• task, subtask, time saved, risk score

olakai activity kpis --json

Core KPIs (executions, compliance, ROI) + custom KPIs

olakai activity kpis --period daily --json

Time-series breakdown

olakai activity kpis --include-atoms --json

Per-event KPI values

olakai agents list --json

Agent metadata

olakai kpis list --json

KPI definitions

olakai activity sessions --agent-id ID --json

Session decoration status + summary counts

Report Type 1: Usage Summary Report

Shows total usage metrics across events, tokens, models, and agents.

Data Collection

Get recent events with analytics

olakai activity list --limit 100 --include-analytics --json > /tmp/events.json

Extract summary metrics

cat /tmp/events.json | jq '{ total_events: (.prompts | length), total_tokens: ([.prompts[].tokens // 0] | add), avg_tokens: ([.prompts[].tokens // 0] | add / length | floor), unique_models: ([.prompts[].model] | unique | length), models: ([.prompts[].model] | group_by(.) | map({model: .[0], count: length})), unique_agents: ([.prompts[].app] | unique | length), agents: ([.prompts[].app] | group_by(.) | map({agent: .[0], count: length})), success_rate: (([.prompts[] | select(.status != "error")] | length) / (.prompts | length) * 100 | floor) }'

Report Template

Usage Summary Report

Generated: [DATE] Period: Last [N] events

Overview

Events by Model

[ASCII BAR CHART]

Events by Agent

[ASCII BAR CHART]

Example Output

Usage Summary Report

Generated: 2025-01-21 Period: Last 100 events

Overview

Events by Model

gpt-4o ████████████████████████████████████ 45 gpt-4o-mini ██████████████████████ 28 claude-3-5 ████████████████ 20 gpt-3.5-turbo █████ 7

Events by Agent

code-assistant ████████████████████████████████ 40 data-analyzer ████████████████████████ 30 chat-support ████████████████████ 25 test-agent ████ 5

Report Type 2: KPI Trends Report

Shows KPI values over time with period-over-period comparisons.

Data Collection

Get KPIs with daily breakdown

olakai activity kpis --period daily --json > /tmp/kpis_daily.json

Get KPIs with weekly breakdown

olakai activity kpis --period weekly --json > /tmp/kpis_weekly.json

Extract trend data

cat /tmp/kpis_daily.json | jq '{ period: "daily", kpis: [.kpis[] | { name: .name, current: .value, trend: .trend, breakdown: .breakdown }] }'

For Custom KPIs with Agent Filter

Get custom KPIs for specific agent

olakai activity kpis --agent-id AGENT_ID --period daily --json | jq '.kpis'

List KPI definitions

olakai kpis list --agent-id AGENT_ID --json | jq '.[] | {name, unit, aggregation}'

Report Template

KPI Trends Report

Generated: [DATE] Agent: [AGENT_NAME] (or "All Agents") Period: [PERIOD]

Core KPIs

Custom KPIs

Daily Trend (Last 7 Days)

[ASCII LINE CHART]

Example Output

KPI Trends Report

Generated: 2025-01-21 Agent: code-assistant Period: Last 7 days

Core KPIs

Custom KPIs

Daily Executions (Last 7 Days)

 150 ┤                           ╭──
 125 ┤              ╭────────────╯
 100 ┤    ╭─────────╯
  75 ┤────╯
  50 ┤
     └──────────────────────────────
      Mon  Tue  Wed  Thu  Fri  Sat  Sun

Report Type 3: Risk Analysis Report

Shows risk distribution, blocked events, and sensitivity patterns.

Data Collection

Get events with risk data

olakai activity list --limit 200 --include-analytics --json > /tmp/events.json

Extract risk metrics

cat /tmp/events.json | jq '{ total_events: (.prompts | length), high_risk: ([.prompts[] | select(.riskScore >= 7)] | length), medium_risk: ([.prompts[] | select(.riskScore >= 4 and .riskScore < 7)] | length), low_risk: ([.prompts[] | select(.riskScore < 4)] | length), blocked: ([.prompts[] | select(.status == "blocked")] | length), blocked_percentage: (([.prompts[] | select(.status == "blocked")] | length) / (.prompts | length) * 100), sensitivity_labels: ([.prompts[].sensitivityLabel] | group_by(.) | map({label: .[0], count: length})), avg_risk_score: ([.prompts[].riskScore // 0] | add / length) }'

Report Template

Risk Analysis Report

Generated: [DATE] Period: Last [N] events

Risk Overview

Risk Distribution

[ASCII BAR CHART]

Events by Sensitivity Label

[ASCII BAR CHART]

High-Risk Event Details (Recent)

Example Output

Risk Analysis Report

Generated: 2025-01-21 Period: Last 200 events

Risk Overview

Risk Distribution

Low (0-3) ████████████████████████████████████████ 172 (86%) Medium (4-6) ████████ 20 (10%) High (7-10) ████ 8 (4%)

Events by Sensitivity Label

Public ████████████████████████████████████ 145 Internal ██████████████████ 42 Confidential ████ 10 Restricted █ 3

High-Risk Events (Recent 5)

Report Type 4: ROI/Efficiency Report

Shows time saved, cost metrics, and productivity gains.

Data Collection

Get KPIs (includes ROI data)

olakai activity kpis --json > /tmp/kpis.json

Get events with time saved data

olakai activity list --limit 100 --include-analytics --json > /tmp/events.json

Extract efficiency metrics

cat /tmp/events.json | jq '{ total_events: (.prompts | length), total_time_saved_minutes: ([.prompts[].timeSavedMinutes // 0] | add), avg_time_saved: ([.prompts[].timeSavedMinutes // 0] | add / length), total_tokens: ([.prompts[].tokens // 0] | add), by_task: ([.prompts[] | select(.task != null)] | group_by(.task) | map({ task: .[0].task, count: length, time_saved: ([.[].timeSavedMinutes // 0] | add) })) }'

Get ROI from KPIs

cat /tmp/kpis.json | jq '.kpis[] | select(.name | contains("ROI") or contains("Compliance"))'

Report Template

ROI/Efficiency Report

Generated: [DATE] Period: Last [N] events

Efficiency Summary

Governance Compliance

Time Saved by Task Type

[ASCII BAR CHART]

ROI Breakdown

[ASCII PIE CHART or TABLE]

Example Output

ROI/Efficiency Report

Generated: 2025-01-21 Period: Last 100 events

Efficiency Summary

Governance Compliance

Time Saved by Task Type

Code Review ████████████████████████████████ 4.2 hrs Bug Analysis ██████████████████████████ 3.5 hrs Documentation ████████████████████ 2.7 hrs Refactoring ████████████████ 2.1 hrs

Productivity Multiplier

Based on avg 7.5 min saved per interaction:

• Daily (50 events): 6.25 hours saved
• Weekly (250 events): 31.25 hours saved
• Monthly (1000 events): 125 hours saved

Report Type 5: Agent Comparison Report

Side-by-side comparison of metrics across multiple agents.

Data Collection

Get all agents

olakai agents list --json > /tmp/agents.json

Get events for comparison

olakai activity list --limit 500 --include-analytics --json > /tmp/events.json

Extract per-agent metrics

cat /tmp/events.json | jq '{ agents: ([.prompts[].app] | unique | map(. as $agent | { name: $agent, events: ([($parent.prompts // [])[] | select(.app == $agent)] | length), tokens: ([($parent.prompts // [])[] | select(.app == $agent) | .tokens // 0] | add), avg_risk: ([($parent.prompts // [])[] | select(.app == $agent) | .riskScore // 0] | add / length) })) }'

Alternative: Get KPIs per agent

for agent_id in $(olakai agents list --json | jq -r '.[].id'); do echo "Agent: $agent_id" olakai activity kpis --agent-id $agent_id --json | jq '.kpis[] | {name, value}' done

Report Template

Agent Comparison Report

Generated: [DATE] Agents Compared: [COUNT]

Activity Volume

KPI Comparison

Risk Profile

[ASCII GROUPED BAR CHART]

Activity Trend by Agent

[ASCII MULTI-LINE CHART]

Example Output

Agent Comparison Report

Generated: 2025-01-21 Agents Compared: 4

Activity Volume

KPI Comparison

Risk Profile by Agent

       Low    Medium    High

code-assist ████████████████████ █ │ 92% 6% 2% data-analyze ██████████████████ ████ ██ 85% 10% 5% chat-support █████████████████████ │ │ 97% 2% 1% test-agent ███████████████████ ██ │ 90% 8% 2%

ASCII Visualization Functions

Bar Chart Generator

To create horizontal bar charts, use this pattern:

Generate bar chart from jq output

cat /tmp/events.json | jq -r ' [.prompts[].model] | group_by(.) | map({model: .[0], count: length}) | sort_by(-.count) | (max_by(.count).count) as $max | .[] | "(.model | .[0:15] | . + " " * (15 - length)) " + ("█" * ((.count / $max * 40) | floor)) + " (.count)" '

Example output:

gpt-4o ████████████████████████████████████████ 45 gpt-4o-mini █████████████████████████ 28 claude-3-5 ██████████████████ 20

Percentage Bar

Show percentage with visual bar

echo "Compliance: ████████████████████░░░░░ 85%"

Pattern:

[LABEL]: [FILLED █ * percentage/4][EMPTY ░ * (25-filled)] [VALUE]%

Trend Indicators

↑ +7% (increase) ↓ -3% (decrease) → 0% (stable)

Quick Reference Commands

Usage Summary

olakai activity list --limit 100 --json | jq '{ events: (.prompts | length), tokens: ([.prompts[].tokens // 0] | add), models: ([.prompts[].model] | unique) }'

KPI Snapshot

olakai activity kpis --json | jq '.kpis[] | {name, value, unit}'

Risk Summary

olakai activity list --limit 100 --json | jq '{ high_risk: ([.prompts[] | select(.riskScore >= 7)] | length), blocked: ([.prompts[] | select(.status == "blocked")] | length) }'

Agent List

olakai agents list --json | jq '.[] | {id, name}'

Per-Agent KPIs

olakai activity kpis --agent-id AGENT_ID --json

Time-Series Data

olakai activity kpis --period daily --json olakai activity kpis --period weekly --json

Generating a Complete Report

Follow this workflow for any report type:

1. Determine scope

AGENT_ID="your-agent-id" # or leave empty for all LIMIT=100

2. Collect data

olakai activity list --limit $LIMIT --include-analytics --json > /tmp/activity.json olakai activity kpis --agent-id $AGENT_ID --json > /tmp/kpis.json olakai agents list --json > /tmp/agents.json

3. Process and format (example for usage summary)

echo "# Usage Summary Report" echo "Generated: $(date +%Y-%m-%d)" echo "" echo "## Overview" cat /tmp/activity.json | jq -r '"| Metric | Value | |--------|-------| | Total Events | (.prompts | length) | | Total Tokens | ([.prompts[].tokens // 0] | add) | | Unique Models | ([.prompts[].model] | unique | length) |"'

Error Handling

No Data Available

Check if events exist

olakai activity list --limit 1 --json | jq '.prompts | length'

If 0, inform user:

"No events found. Ensure your agent is sending events to Olakai."

Agent Not Found

Verify agent exists

olakai agents list --json | jq '.[] | select(.id == "AGENT_ID")'

If empty, list available agents:

olakai agents list --json | jq '.[] | {id, name}'

Missing Permissions

Re-authenticate if needed

olakai logout && olakai login olakai whoami # Verify

Best Practices

• Always use --json flag for programmatic processing

• Pipe through jq for clean data extraction

• Cache data locally when generating multi-section reports

• Include timestamps in all reports

• Show data freshness - how recent the events are

• Handle empty states gracefully with informative messages