Monocle
Monocle2AIOpenTelemetry-native tracing SDK for GenAI applications and agents, built under the Linux Foundation AI & Data umbrella. It ships a metamodel mapping agents, prompts, tools and vector operations to standardized OTLP span attributes, framework adapters for LangChain, LlamaIndex, Haystack and Google ADK, and a companion trace-assertion testing library, monocle-test-tools. Traces export as OTLP to any collector or backend, with no hosted service of its own.
Its PyPI homepage points back at monocle2ai/monocle; the publisher of record is Okahu, the company that contributed Monocle to the Linux Foundation and maintains it, so a different publisher name is not a different project.
Openness
5 high confidence- license
- Apache-2.0(OSI, single root LICENSE)
- source
- public
- core features withheld
- no — a Linux Foundation AI & Data project with no hosted platform or commercial tier of any kind
Monocle is a pure OpenTelemetry instrumentation SDK with a companion test-assertion library. Everything the product does - the metamodel, the framework adapters, the test tools - ships in this Apache-2.0 repository, and there is no vendor selling a hosted or gated tier alongside it.
- https://api.github.com/repos/monocle2ai/monocle/git/trees/main?recursive=1 recorded 2026-09-02
The whole repository tree, 873 paths. No ee/, enterprise, premium or entitlement path exists anywhere; the only LICENSE is the root Apache-2.0 file.
- https://raw.githubusercontent.com/monocle2ai/monocle/main/LICENSE recorded 2026-09-02
Apache License, Version 2.0 full text, the sole license file in the repository.
- https://raw.githubusercontent.com/monocle2ai/monocle/main/README.md recorded 2026-09-02
Built under the Linux Foundation AI & Data umbrella. "This repository contains the Python implementation of Monocle's tracing SDK and metamodel", plus a separate monocle-test-tools package, with exports to local JSON files, console or an OTLP-compatible backend - no hosted service is described anywhere.
Adoption
2 medium confidenceAdoption is measured by monthly PyPI downloads of monocle_apptrace, the SDK itself, since installing the package is installing Monocle and there is no separate platform behind it. GitHub stars on the project, which started in 2024, corroborate a still-growing community without adding to the reading.
- https://api.github.com/repos/monocle2ai/monocle recorded 2026-09-02
stargazers_count = 337, forks_count = 58, created_at 2024-06-15, pushed_at 2026-09-01 (actively maintained)
- https://pypi.org/pypi/monocle_apptrace/json recorded 2026-09-02
info.name monocle-apptrace, version 0.8.14, project_urls Homepage https://github.com/monocle2ai/monocle, author "Okahu Inc." <okahu-pypi@okahu.ai>, license Apache-2.0 - the package points back at the declared repository.
- https://pypistats.org/api/packages/monocle_apptrace/recent recorded 2026-09-02
last_month = 48,842 downloads for monocle-apptrace (last_week 12,339, last_day 2,256)
Capability
3 high confidenceMonocle is a standards-focused tracing and testing layer: broad framework coverage across LangChain, LlamaIndex, Haystack, Google ADK and more, plus a distinctive trace-assertion test tool, monocle-test-tools. It stops at instrumentation and testing, with no prompt management, evaluation, dataset or playground surface - the same shape as OpenLLMetry in this category and a step below Langfuse.
- https://raw.githubusercontent.com/monocle2ai/monocle/main/README.md recorded 2026-09-02
Feature description: a metamodel for entities (agents, prompts, responses, tools, vector operations), instrumentation adapters for LangChain, LlamaIndex, Haystack, Google ADK, OpenAI and Anthropic, OTLP-compatible export, and a separate monocle-test-tools package for asserting on traces themselves - no evaluation, prompt-management or dataset section exists.
Unchanged since 2026-10-07 (last edited, not re-checked)