Aigon collects telemetry from every agent session — Claude Code, Gemini CLI, and Codex CLI — and normalizes it into a unified format. This gives you a single view of token usage, cost, and activity across your entire multi-agent workflow.

What is collected

After each feature session, Aigon parses agent-specific transcript files to extract:

Telemetry is stored locally in .aigon/telemetry/ — nothing is sent to external services.

Agent parsers

Each agent CLI writes transcripts in a different format. Aigon includes a parser for each:

Parsers run automatically during feature-close and feature-submit — no manual step is needed.

Activity breakdown

Every telemetry record includes an activity field that classifies the session:

This distinction lets you answer questions like:

• How much of my token spend goes to implementation vs. evaluation?
• Which activity type is most expensive per feature?
• Are review sessions getting shorter over time?

Cost attribution

Aigon maintains a pricing table covering Claude, Gemini, and GPT-5 model families. Costs are computed per-turn using actual token counts:

cost = (input_tokens × input_rate)
     + (output_tokens × output_rate)
     + (cache_read_tokens × input_rate × 0.10)
     + (cache_write_tokens × input_rate × 1.25)

Cache-aware pricing means you see the real cost, not a worst-case estimate. If a model ID isn’t in the pricing table, Aigon falls back to Sonnet-tier rates as a conservative default.

Reading the Reports tab

The Dashboard Reports tab visualizes telemetry data across all your repos. Here’s how to find the telemetry-specific views:

Summary cards

The Summary sub-tab shows aggregate metrics. Look for:

• Features Completed with cost attribution per agent
• Agent Leaderboard showing per-agent token spend alongside performance metrics

Charts

The Charts sub-tab includes time-series charts with a daily/weekly/monthly granularity toggle and synchronized panning. Two charts are available on the free tier:

• Features Completed — bar chart of features closed per period
• Commits — stacked bar chart (feature vs. non-feature commits)

With Aigon Pro, four additional charts unlock:

• Median Cycle Time — bar chart of median cycle time per period
• Commits per Feature — median commits per feature completed
• Rework Ratio — percentage of fix commits per period
• Tokens Used — total billable tokens consumed per period

The Pro Insights tab adds even deeper token analytics — per-agent cost breakdowns, token efficiency trends, and token usage by activity type. See the Insights guide for details.

Free tier vs. Pro

The free Reports tab gives you:

• Two charts (Features Completed, Commits)
• Summary metric cards (features completed, commits, lines changed)
• Detail tables (features and commits)

Aigon Pro adds to the Reports tab:

• Four additional charts (Cycle Time, CPF, Rework Ratio, Tokens Used)
• Additional summary cards (Commits/Feature, Rework Ratio)
• Agent leaderboard

Pro also adds the Insights tab with amplification analytics:

• First-pass rate, median cost/feature, token efficiency metrics
• Cost and token-efficiency sparkline trends
• Token analytics (distribution, per-agent breakdown, activity breakdown)
• AI-generated insights and coaching recommendations

See the Amplification Dashboard guide for details.

Troubleshooting

No telemetry data showing up?

1. Make sure you’ve closed at least one feature — telemetry is collected during feature-close
2. Check that transcript files exist for your agent (see the parser table above)
3. Verify the agent session ran long enough to generate transcript data

Cost showing as $0?

The model ID in the transcript may not match the pricing table. Aigon falls back to Sonnet pricing, but if the transcript has no usage blocks at all, costs will be zero. This typically means the agent session was too short to generate meaningful output.

Next steps

• Dashboard guide — full walkthrough of all dashboard tabs
• Insights (Amplification) — Pro-tier quality and token analytics
• Fleet Mode — multi-agent workflows that generate the most telemetry data