OpenAI Codex

OpenAI Codex has native OpenTelemetry support. Once enabled, Oodle collects metrics and event logs so you can track token usage, session activity, tool calls, and WebSocket request patterns across your organization.

Getting Started

1. Enable Telemetry

The fastest way is to use the integration tile in the Oodle UI:

1. Navigate to Settings → Integrations
2. Open the AI Cost Management section
3. Click the Codex Observability tile
4. Select an API key and follow the steps shown

Alternatively, add the following to your Codex configuration file (~/.codex/config.toml):

#:schema https://developers.openai.com/codex/config-schema.json

[otel]
environment = "production"
log_user_prompt = true

exporter = { otlp-http = {
  endpoint = "https://<LOGS_ENDPOINT>/ingest/otel/v1/logs",
  protocol = "binary",
  headers = {
    "X-API-KEY" = "<API_KEY>",
    "X-OODLE-INSTANCE" = "<INSTANCE_ID>",
  },
}}

metrics_exporter = { otlp-http = {
  endpoint = "https://<METRICS_ENDPOINT>/v2/otlp/metrics/<INSTANCE_ID>",
  protocol = "binary",
  headers = {
    "X-API-KEY" = "<API_KEY>",
    "X-OODLE-INSTANCE" = "<INSTANCE_ID>",
  },
}}

Replace <METRICS_ENDPOINT>, <API_KEY>, and <INSTANCE_ID> with values from the integration tile.

tip

The integration tile in the Oodle UI generates a ready-to-copy config.toml with the correct endpoints and API key pre-filled.

Alternative: Environment Variables

For quick testing you can export the standard OpenTelemetry variables directly in your shell:

export OTEL_EXPORTER_OTLP_PROTOCOL=http/protobuf
export OTEL_EXPORTER_OTLP_METRICS_ENDPOINT=https://<METRICS_ENDPOINT>/v2/otlp/metrics/<INSTANCE_ID>
export OTEL_EXPORTER_OTLP_LOGS_ENDPOINT=https://<LOGS_ENDPOINT>/ingest/otel/v1/logs
export OTEL_EXPORTER_OTLP_HEADERS="X-API-KEY=<API_KEY>, X-OODLE-INSTANCE=<INSTANCE_ID>"
export OTEL_LOG_USER_PROMPTS=1
export OTEL_METRIC_EXPORT_INTERVAL=10000
export OTEL_LOGS_EXPORT_INTERVAL=5000

2. Roll Out to Your Team

Commit the config.toml above to your dotfiles repository or place it in each developer's ~/.codex/ directory via your configuration management tool.

3. Verify Data

Once telemetry starts flowing, navigate to AI Assistants → Codex in the Oodle sidebar.

Charts Dashboard

The Charts tab embeds a Grafana dashboard with panels covering:

Panel	Description
Total Tokens	Aggregate token usage over the selected range
Conversations	Number of Codex conversations started
WebSocket Requests	Total WebSocket requests made
Tool Calls	Total tool invocations
Token Usage	Time series by token type (input, output, cached, reasoning)
Conversations Over Time	Conversation start trends
Tool Call Activity	Tool invocations over time
WebSocket Requests	Success vs failure rates
Avg Turn E2E Duration	End-to-end latency per turn
Avg TTFT	Time to first token
Avg TTFM	Time to first message
Avg WS Event Duration	WebSocket event processing time
Startup Prewarm Duration	Agent startup prewarm latency
Shell Snapshot Duration	Shell snapshot capture latency

Sessions

The Sessions tab shows individual Codex sessions:

Column	Description
Start Time	When the session began
Session	Codex conversation ID
User	Email of the developer
Surface	Where Codex ran — app-server (IDE/extension) or cli (terminal)
Duration	Wall-clock duration
Tokens	Total tokens (input + output + cached)

note

Codex does not currently export cost data, so the Cost column is hidden for Codex sessions. Duration is blank for a session until it has reported more than one data point.

Click any row to open a Session Detail drawer showing a turn-by-turn timeline of every event.

Session Detail Drawer

The drawer displays:

• Session metadata — user, model, app version
• Aggregated stats — tokens, tool calls, errors, duration (only populated fields are shown)
• Turn-by-turn timeline — each turn is collapsible and shows individual events (SSE events, WebSocket events, tool calls). Every event row is expandable to reveal the full raw JSON payload.

Tool Usage

The Tool Usage tab embeds a Codex-specific Grafana dashboard (separate from the Claude Code one) driven by the log-derived oodle_logs_codex_tool_* metrics:

Panel	Description
Tool Calls Over Time	Invocations per tool name
Top Tools by Call Count	Most-used tools in the range
Avg Tool Duration	Mean execution time per tool
Tool Outcomes	Success vs failure call counts
MCP Tool Calls by Server	MCP invocations grouped by server

Filter by Surface, Model, and User from the dashboard template variables.

Token Usage

The Token Usage tab breaks down token consumption from the log-derived oodle_logs_codex_token_usage metric:

Panel	Description
Token Usage by User	Daily tokens per developer
Token Usage by Model	Daily tokens per model
Token Usage by Token Type	input / output / cached_input / reasoning_output
Token Usage by Surface	app-server vs CLI
Token Usage by App Version	Daily tokens per Codex version
Top Sessions by Token Usage	Highest-token sessions in range

Cached-input and reasoning-output tokens are tracked separately from input/output — use the Token Type filter to isolate a category.

What Gets Collected

Metrics

Codex exports the following as OpenTelemetry metrics (delta temporality):

Metric	Labels	Description
codex_turn_token_usage	model, token_type, originator	Token count by type (input, output, cached, reasoning)
codex_thread_started	model, originator	Conversations started
codex_turn_tool_call	model, originator	Tool invocations
codex_websocket_request	model, success	WebSocket API requests
codex_turn_e2e_duration_ms	model	End-to-end turn latency
codex_turn_ttft_ms	model	Time to first token
codex_turn_ttfm_ms	model	Time to first message
codex_ws_event_duration_ms	model	WebSocket event processing time
codex_startup_prewarm_duration_ms	—	Startup prewarm latency
codex_shell_snapshot_duration_ms	—	Shell snapshot capture latency

Events (Logs)

Events are exported via the OpenTelemetry logs protocol. Each event has an attributes.event.name field:

Event Type	Key Attributes
codex.conversation_starts	Conversation ID, model, user email
codex.sse_event	Model, event kind, token counts, duration
codex.websocket_event	Model, event kind, duration, success
codex.websocket_request	Model, duration, success
codex.websocket_connect	Model, duration
codex.user_prompt	Prompt text, prompt length
codex.tool_decision	Tool name, decision, source
codex.tool_result	Tool name, success, duration, error

Events are grouped by conversation.id to reconstruct the turn-by-turn timeline on the Sessions tab.

Further Reading

• Codex Observability Documentation
• OpenTelemetry Integration