CtrlK
BlogDocsLog inGet started
Tessl Logo

pantheon-ai/promql-toolkit

Complete PromQL toolkit with generation and validation capabilities

94

Quality

94%

Does it follow best practices?

Impact

Pending

No eval scenarios have been run

SecuritybySnyk

Advisory

Suggest reviewing before use

Overview
Quality
Evals
Security
Files

task.mdgenerator/evals/scenario-0/

Build a RED Method Dashboard Query Set for an API Gateway

Problem/Feature Description

A platform engineering team is building a Grafana dashboard for their API gateway (api-gateway). They want to implement the RED method (Rate, Errors, Duration) across all endpoints. The team's Prometheus scrapes two metric types:

  • http_requests_total — a counter with labels job, endpoint, and status_code
  • node_memory_MemAvailable_bytes — a gauge (available memory in bytes) with label instance

An intern recently submitted dashboard panels that apply the same rate() function to both metrics. The senior engineer suspects this is wrong for the gauge metric and has asked you to generate a correct set of queries for the dashboard.

The team wants three queries:

  1. Per-endpoint request throughput (requests per second over a 5-minute window), broken down by endpoint
  2. The overall error ratio (5xx responses as a fraction of all requests) over 5 minutes for the api-gateway job
  3. Available memory on each instance (current value, with instance label preserved)

Produce the queries in a file called queries.promql. Each query should be on its own line with a comment above it explaining what it measures.

Output Specification

Produce a file queries.promql containing the three queries described above, each preceded by a # comment describing what it measures.

generator

evals

scenario-0

capability.txt

criteria.json

task.md

instructions.json

summary_infeasible.json

summary.json

SKILL.md

tile.json