godot-runtime-telemetry-lab 0.1.6

Summarize, adapt, and compare Godot runtime telemetry with rule metadata for CI reports.

Godot Runtime Telemetry Lab

godot-runtime-telemetry-lab summarizes and compares lightweight runtime telemetry from Godot scenario runs, smoke tests, soak tests, or project-owned debug exporters. It is designed for CI artifacts and release reviews, not as a replacement for Godot's built-in profiler.

Install

python -m pip install godot-runtime-telemetry-lab

From a source checkout:

python -m pip install -e .\godot-runtime-telemetry-lab

Quick Start

godot-telemetry-lab summarize reports\runtime --format markdown --output reports\runtime.md
godot-telemetry-lab compare reports\baseline reports\current --format json --output reports\runtime-compare.json
godot-telemetry-lab timeline reports\runtime --format html --output reports\runtime-timeline.html
godot-telemetry-lab adapt reports\godot-monitor.csv --format json --output reports\runtime-normalized.json
godot-telemetry-lab adapt godot-runtime-telemetry-lab\examples\tiny-godot-monitor\godot-performance-monitor.csv --format json --output reports\runtime-normalized.json
godot-telemetry-lab budget init --profile android-high --output reports\runtime-budget.json

Input Shape

The tool accepts .json or .csv files. JSON files can contain a list of samples, or an object with a samples, frames, or events list. Commands also accept a directory containing .json and .csv files. If the path is wrong, the file extension is unsupported, the directory has no telemetry files, or input cannot be read, the command emits a normal finding with next-step guidance instead of a Python traceback.

{
  "samples": [
    {"scenario": "menu", "frame_ms": 12.4, "physics_ms": 2.1, "memory_mb": 180},
    {"scenario": "menu", "frame_ms": 18.8, "physics_ms": 2.5, "memory_mb": 181}
  ]
}

Recognized numeric fields are frame_ms, physics_ms, memory_mb, nodes, and draw_calls. Timeline output also uses optional time_s, timestamp_s, frame, scenario, phase, and event fields when they are present.

Use adapt when a project-owned Godot exporter writes monitor names such as fps, Performance.MEMORY_STATIC, Performance.OBJECT_NODE_COUNT, or Performance.RENDER_TOTAL_DRAW_CALLS_IN_FRAME. The command emits normalized samples that can be fed back into summarize, compare, or timeline. For official Godot Performance monitors, adapt treats Performance.TIME_PROCESS and Performance.TIME_PHYSICS_PROCESS as seconds and converts them to milliseconds. Memory monitors such as Performance.MEMORY_STATIC, Performance.RENDER_VIDEO_MEM_USED, Performance.RENDER_TEXTURE_MEM_USED, and Performance.RENDER_BUFFER_MEM_USED are treated as bytes and converted to MiB.

Commands

• summarize: reports sample counts, frame percentiles, and budget findings.
• compare: compares current telemetry with a baseline and reports frame p95 and memory max regressions.
• timeline: renders a frame and memory timeline as HTML, SVG, Markdown, text, or JSON.
• adapt: normalizes common Godot monitor and debug-exporter field names.
• budget init: writes starter budget JSON for desktop-dev, android-high, android-low, or html5.

Budget files can be reused with summarize, compare, and timeline:

godot-telemetry-lab budget init --profile android-high --output reports\runtime-budget.json
godot-telemetry-lab timeline reports\runtime --budget-file reports\runtime-budget.json --format html --output reports\runtime-timeline.html

compare includes frame_p95_delta_ms and memory_delta_mb in JSON reports, so CI jobs and dashboards can show the size of a runtime change without parsing the full baseline and current summaries. JSON reports also include metadata.rules, a compact catalog of runtime telemetry rule titles and help text for scripts, dashboards, and CI comments.

Outputs

• text: local terminal report.
• json: CI and scripts.
• markdown: PR comments and release notes.
• html: self-contained timeline report.
• svg: embeddable timeline chart.

Example

The package includes a tiny runtime fixture:

godot-telemetry-lab timeline godot-runtime-telemetry-lab\examples\tiny-runtime-run --format html --output reports\runtime-timeline.html

It also includes a tiny Godot monitor CSV fixture:

godot-telemetry-lab adapt godot-runtime-telemetry-lab\examples\tiny-godot-monitor\godot-performance-monitor.csv --format json --output reports\runtime-normalized.json