Unlock Full Agentic AI Tokenomics
Inside GitHub Copilot

STEP 1 OF 4

Download update-tokenomics.py into your .copilot/instructions folder

Run this one-liner in PowerShell (Windows) or Terminal (macOS/Linux) to save the script in your user directory:

New-Item -ItemType Directory -Force "C:\Users\YourAliasName\.copilot\instructions" | Out-Null
Invoke-WebRequest -Uri "https://raw.githubusercontent.com/ShahabName/GetTokenomics/main/update-tokenomics.py" `
  -OutFile "C:\Users\YourAliasName\.copilot\instructions\update-tokenomics.py"

STEP 2 OF 4

Add the Modular Tokenomics Block & ON/OFF Switch to global.instructions.md

Open C:\Users\YourAliasName\.copilot\instructions\global.instructions.md in VS Code and paste the cleaned modular snippet below:

## Record and report daily tokenomics at the end of execution

- **Tokenomics Tracking Switch**: `TOKENOMICS_SWITCH: ON` *(Change to `OFF` to disable tokenomics CSV logging and response reporting; change to `ON` to enable)*
- **Switch Rule**: Before recording or reporting tokenomics, check `TOKENOMICS_SWITCH` above. If it is set to `OFF`, skip running `update-tokenomics.py`, do not create or modify `tokenomics-CCYYMMDD.csv`, and omit the `## Tokenomics` section from the response.

When `TOKENOMICS_SWITCH: ON`, at the end of every prompt or task execution, record the prompt's tokenomics and AIC/AIU consumption in today's daily CSV file in the workspace root (`tokenomics-CCYYMMDD.csv`, where `CCYYMMDD` is the current date in `YYYYMMDD` format, such as `tokenomics-20261024.csv`) and display a summary in the final response.

- Check whether `tokenomics-CCYYMMDD.csv` exists in the current workspace root directory with the following 29 header columns:
  ```csv
  timestamp_utc,session_id,challenge_or_task,turn_index,sequence_id,initiator,router_model,source_model,model_display_name,reasoning_effort,api_calls_count,user_prompt_multiplier,total_request_multiplier_cost,total_input_tokens,uncached_input_tokens,cache_read_tokens,cache_write_tokens,output_tokens,reasoning_tokens,tool_schema_tokens,cache_hit_ratio_pct,total_nano_aiu,total_aiu_credits,duration_ms,ttft_ms,tools_invoked,finish_reason,ai_credits_status,prompt_summary
  ```
- If `tokenomics-CCYYMMDD.csv` does not exist yet, create it with the 29 header columns above.
- If `tokenomics-CCYYMMDD.csv` already exists, append the latest prompt turn tokenomics data from `C:\Users\YourAliasName\.copilot\session-state\<session-id>\events.jsonl` (and refresh any previous turn in the session whose final tool-call usage settled after the prior turn). You can run `python "C:\Users\YourAliasName\.copilot\instructions\update-tokenomics.py"` to perform this check, creation, and append automatically.
- At the end of every prompt response (immediately above `## Files`), include a `## Tokenomics` section showing the tokenomics usage data recorded in `tokenomics-CCYYMMDD.csv` (model, API calls, input/cache/output/reasoning tokens, cache hit %, total NanoAIU, total AIU/AIC credits, duration, and a clickable link to `[tokenomics-CCYYMMDD.csv](C:/full/absolute/path/to/tokenomics-CCYYMMDD.csv)` using its absolute filesystem path with `/` separators).

STEP 3 OF 4

Run Any Prompt in GitHub Copilot Chat and See Live Tokenomics

Every response now automatically updates today's tokenomics-CCYYMMDD.csv and displays the structured ## Tokenomics summary right above clickable ## Files links:

STEP 4 OF 4

Toggle ON/OFF Anytime or Analyze Your 29-Column CSV

Use the CLI flags or edit TOKENOMICS_SWITCH: OFF whenever you want to pause tracking:

# Check current status (ON or OFF)
python "C:\Users\YourAliasName\.copilot\instructions\update-tokenomics.py" --status

# Turn OFF tokenomics tracking
python "C:\Users\YourAliasName\.copilot\instructions\update-tokenomics.py" --off

# Turn ON tokenomics tracking
python "C:\Users\YourAliasName\.copilot\instructions\update-tokenomics.py" --on