llm-observe-proxy 0.1.1

OpenAI-compatible LLM proxy with SQLite request capture, observability, and an admin UI.

LLM Observe Proxy

llm-observe-proxy is an OpenAI-compatible, record-only proxy for inspecting LLM traffic. It forwards requests to an upstream /v1 API, stores requests and responses in SQLite, and provides a polished local admin UI for browsing, pretty-printing, trimming, and changing runtime settings.

It is useful when you want LiteLLM-style observability without introducing a full gateway or external database.

Published package: https://pypi.org/project/llm-observe-proxy/

Features

• OpenAI-compatible passthrough route: ANY /v1/{path:path}.
• SQLite capture for request/response headers, bodies, status, timing, model, endpoint, streaming state, tool-call signals, image assets, and errors.
• Admin UI for searching and browsing captured traffic.
• Detail pages with response render modes for JSON, plain text, Markdown, tool calls, and raw SSE streams.
• Request image gallery for data URL and remote image references.
• Settings UI for upstream URL, incoming host/port preferences, all-IPs exposure, and retention trimming.
• No authentication by default, intended for local or trusted development networks.

Install

From PyPI with pip:

python -m pip install llm-observe-proxy
llm-observe-proxy

From PyPI with uv:

uv tool install llm-observe-proxy
llm-observe-proxy

Run it once without installing:

uvx llm-observe-proxy

For local development from this repository:

C:\Python\Python313\python.exe -m venv .venv
.\.venv\Scripts\python.exe -m pip install -e .[dev]
.\.venv\Scripts\llm-observe-proxy.exe

By default, the proxy listens on:

http://localhost:8080

and forwards requests to:

http://localhost:8000/v1

Open the admin UI:

http://localhost:8080/admin

Usage

Point an OpenAI-compatible client at the proxy:

from openai import OpenAI

client = OpenAI(
    api_key="local-dev-key",
    base_url="http://localhost:8080/v1",
)

response = client.chat.completions.create(
    model="gpt-demo",
    messages=[{"role": "user", "content": "Hello through the proxy"}],
)
print(response.choices[0].message.content)

Run on a different port:

llm-observe-proxy --port 8090

Expose on all interfaces:

llm-observe-proxy --expose-all-ips

Set the upstream from the CLI:

llm-observe-proxy --upstream-url http://localhost:8000/v1

You can also change the upstream URL and next-start incoming host/port settings from /admin/settings.

Screenshots

Screenshots are generated from a seeded demo database and stored in docs/screenshots.

Request browser	Tool calls

Images	Settings

Additional screenshots:

• Simple request detail
• Streaming SSE detail

Regenerate screenshots:

.\.venv\Scripts\python.exe scripts\seed_demo_db.py .tmp\screenshots.sqlite3
.\.venv\Scripts\python.exe scripts\capture_screenshots.py --database .tmp\screenshots.sqlite3 --output docs\screenshots

Routes

• ANY /v1/{path:path}: OpenAI-compatible pass-through proxy.
• GET /admin: request browser.
• GET /admin/requests/{id}: request/response detail view.
• GET /admin/settings: upstream settings and retention tools.
• POST /admin/settings/incoming: update incoming host/port settings for next startup.
• POST /admin/settings/upstream: update upstream URL.
• POST /admin/trim: delete records older than N days.
• GET /healthz: health check.

Configuration

Environment variables:

Variable	Default	Purpose
LLM_OBSERVE_DATABASE_URL	sqlite:///./llm_observe_proxy.sqlite3	SQLite SQLAlchemy URL.
LLM_OBSERVE_INCOMING_HOST	localhost	Bind host when not exposing all IPs.
LLM_OBSERVE_INCOMING_PORT	8080	Bind port.
LLM_OBSERVE_EXPOSE_ALL_IPS	false	Bind to 0.0.0.0 when true.
LLM_OBSERVE_UPSTREAM_URL	http://localhost:8000/v1	Upstream OpenAI-compatible /v1 base URL.
LLM_OBSERVE_LOG_LEVEL	INFO	Uvicorn log level.

Incoming host/port settings saved in the UI are used on the next process startup; they do not rebind a currently running process.

Tests

.\.venv\Scripts\ruff.exe check src tests
.\.venv\Scripts\python.exe -m compileall -q src tests
.\.venv\Scripts\pytest.exe -q

The test suite starts a fake upstream on localhost:8080/v1, so stop any local process using port 8080 before running tests. See docs/tests/README.md for the full coverage matrix.

Publishing

See docs/publishing.md for name checks, build commands, and the pre-publish checklist.

License

MIT. See LICENSE.