tts-adoption-report

Generates daily TTS adoption reports from ClickHouse and Datadog metrics posted to Slack.

7|12|Updated Mar 26, 2026
One-click install
npx skills add https://github.com/OpenRouterTeam/docs --skill tts-adoption-report-openrouterteam
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: tts-adoption-report
Source: https://github.com/OpenRouterTeam/docs/tree/main/.agents/skills/tts-adoption-report
Command: npx skills add https://github.com/OpenRouterTeam/docs --skill tts-adoption-report-openrouterteam

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Tracking text-to-speech model adoption requires manually combining usage metrics from ClickHouse with failure logs from Datadog and formatting them into a shareable report, which is tedious and error-prone when done daily. ## Core Features & Use Cases - Automated usage analytics: Queries ClickHouse for weekly/daily unique users (new vs. repeat), cost, revenue, duration percentiles, and model mix for TTS generations. - Reliability reporting: Combines ClickHouse successes with Datadog failure logs from cfw-tts-api, categorizing failures (provider-fault, worker-fault, policy-block, budget/limit) so the headline failure rate reflects real reliability. - Slack delivery: Posts one threaded summary to #metrics-multi-modality with day-aware framing (week-over-week on Mondays, day-over-day Tuesday–Friday), plus a PostHog rankings SEO engagement section. - Use Case: A product team runs this on a weekday schedule to monitor TTS adoption since launch, catching model-mix shifts and failure spikes without writing queries by hand. ## Quick Start Run the TTS adoption report for yesterday and post the summary to the metrics-multi-modality Slack channel.

Frequently Asked Questions about tts-adoption-report

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a daily TTS adoption report?▼

Run the skill on a weekday schedule; it queries ClickHouse default.generations filtered by api_type='tts' for users, revenue, and model mix, then queries Datadog for cfw-tts-api failures, and posts a threaded summary to Slack channel C0BCV3YNZKK.

How do I combine ClickHouse and Datadog for API reliability metrics?▼

Count successes from ClickHouse default.generations and failures from Datadog logs matching the one-per-failure event. Categorize failures by error_location and compute the headline failure rate using only provider-fault and worker-fault categories.

Why does the report show duration metrics unavailable?▼

The generation_time column is NULL on TTS rows in default.generations because cfw-tts-api does not populate it. The report emits an explicit unavailable line instead of fabricated zeros until the column becomes populated.

How do I handle Datadog get_logs pagination limits?▼

The get_logs API caps results at 1000 per request. Split the time window into sub-windows and recursively re-query any window returning 1000 or more results, then sum the leaf counts while throttling calls to avoid rate limits.

Can I run the TTS report for a specific past date?▼

Yes, manual runs support a date override to report on a specific day instead of yesterday, plus a Slack channel override. By default the report covers the most recent complete UTC day.