Releases·Tool Release
Dev releases boothpy v0.5.0-beta, an open-source Python LLM checkpoint library, and seeks developers to stress-test it
An independent developer released v0.5.0-beta of boothpy (BOOTH), an open-source Python library providing a zero-dependency, provider-agnostic checkpoint layer between LLM calls and applications (LLM → BOOTH → accept/retry/reject). The release post recruits developers to test the library across providers like OpenAI, Anthropic, Groq, Mistral, and Ollama, with a GitHub issue open for provider integration examples and benchmarks. It matters to practitioners building reliability tooling who want an early-stage, MIT-licensed option to evaluate.
6 engagement2 sourcesWed, Sep 23, 2026, 06:20 AM
AI Overview
- boothpy (BOOTH) is a zero-dependency, provider-agnostic Python layer that checkpoints LLM calls as accept, retry, or reject
- Version 0.5.0-beta is MIT-licensed, installable from PyPI as boothpy, and works with OpenAI, Anthropic, Groq, Mistral, and Ollama
- The developer asks testers to benchmark calls with and without BOOTH via GitHub Issue #2, 'Add an LLM Provider Integration Example & Benchmark'
Related
Products·launched
BOOTH: Python checkpoint library for validating LLM outputs before they reach your application
BOOTH is a lightweight, dependency-free Python library that validates LLM outputs before they reach application code. It offers check()/acheck() for ambiguity and confidence checks and check_with_evidence() to compare answers against RAG-retrieved evidence. It targets developers building LLM-powered applications who need output reliability guardrails.
BOOTH is a lightweight, dependency-free Python library that validates LLM outputs before they reach application code. It offers check()/acheck() for ambiguity and confidence checks and check_with_evidence() to compare answers against RAG-retrieved evidence. It targets developers building LLM-powered applications who need output reliability guardrails.
1 engagement1 sourceredditFri, Oct 2, 2026, 10:20 PM
Releases·Tool Release
I built a small Python library to add retries, caching, fallbacks, budgets, and guardrails around native LLM SDK calls
A developer open-sourced callm, a Python library that decorates existing OpenAI, Anthropic, and Google GenAI SDK call sites with retries, caching, fallbacks, cost budgets, and guardrails. Unlike LiteLLM or a gateway, it keeps native SDK calls intact and runs safeguards in-process. It is early-stage (v0.1.0) and the author is soliciting production feedback.
A developer open-sourced callm, a Python library that decorates existing OpenAI, Anthropic, and Google GenAI SDK call sites with retries, caching, fallbacks, cost budgets, and guardrails. Unlike LiteLLM or a gateway, it keeps native SDK calls intact and runs safeguards in-process. It is early-stage (v0.1.0) and the author is soliciting production feedback.
1 engagement1 sourceredditWed, Sep 16, 2026, 01:35 PM
Products·building
RuntimeBrain: Python layer explaining OpenAI calls before execution
RuntimeBrain is a small Python layer around OpenAI API calls that provides a pre-call explanation of the execution decision, including what the request is, why it's being made, and potential risks. It aims to help developers understand and review their LLM calls without altering the response or adding extra LLM calls. The project is currently seeking testers.
RuntimeBrain is a small Python layer around OpenAI API calls that provides a pre-call explanation of the execution decision, including what the request is, why it's being made, and potential risks. It aims to help developers understand and review their LLM calls without altering the response or adding extra LLM calls. The project is currently seeking testers.
1 engagement1 sourceredditMon, Aug 24, 2026, 06:59 PM
Products·building
Auscope: Real-time LLM output verification tool
Auscope is a tool that audits LLM responses in production to detect when the model is confidently wrong. It uses three independent models from different providers to check each response, with a fourth 'chairman' model resolving disagreements, providing a verdict of verified, uncertain, or unreliable. It integrates via a one-line Python client and supports multiple LLM providers.
Auscope is a tool that audits LLM responses in production to detect when the model is confidently wrong. It uses three independent models from different providers to check each response, with a fourth 'chairman' model resolving disagreements, providing a verdict of verified, uncertain, or unreliable. It integrates via a one-line Python client and supports multiple LLM providers.
8 engagement1 sourceredditMon, Aug 10, 2026, 04:11 PM
Releases·Tool Release
Developer releases pytest-torch, an alpha-stage pytest plugin for testing PyTorch models
A developer released pytest-torch, an open-source pytest plugin in alpha that reduces repetitive PyTorch test boilerplate like device setup, reproducibility checks, and output comparisons. It bundles inference checks, compile/export parity tests, checkpoint tests, accuracy evaluation, and layer-by-layer divergence diagnostics. The author is soliciting feedback and examples of hard-to-test model bugs.
A developer released pytest-torch, an open-source pytest plugin in alpha that reduces repetitive PyTorch test boilerplate like device setup, reproducibility checks, and output comparisons. It bundles inference checks, compile/export parity tests, checkpoint tests, accuracy evaluation, and layer-by-layer divergence diagnostics. The author is soliciting feedback and examples of hard-to-test model bugs.
3 engagement1 sourceredditTue, Oct 13, 2026, 11:42 PM