Public catalog
Browse tools
14 shown. Search ranks against problem language; browsing defaults to stronger public signals.
dspy
optimize LLM prompts and pipelines programmatically
Verified 2026-08-10 · 12,108 stars
safe-github-onboarding
safe patterns for agents onboarding to GitHub repos
Verified 2026-08-10 · 6,877 stars
langfuse
observe LLM traces in production
Verified 2026-08-10 · 6,855 stars
persona-lock-eval
evaluate whether an agent keeps its persona locked
Verified 2026-08-10 · 5,452 stars
promptfoo
regression test LLM prompts and RAG outputs in CI
Verified 2026-08-10 · 5,452 stars
prove-it-side-effect-check
require agents to prove side effects before claiming done
Verified 2026-08-10 · 5,452 stars
last30days-skill
skill to pull last 30 days of signals for research
Verified 2026-08-10 · 5,216 stars
calc-outside-llm
do exact math outside the LLM to avoid arithmetic errors
Verified 2026-08-10 · 5,099 stars
deepeval
pytest style LLM evaluation metrics
Verified 2026-08-10 · 3,114 stars
stripe-node
use stripe-node for browser automation
Verified 2026-08-10 · 683 stars
printing-press-library
compose and package agent skills without folder chaos
Verified 2026-08-10 · 167 stars
browser-tools-mcp
use browser-tools-mcp for browser automation
Verified 2026-08-10
pyrefly
type-check a large Python codebase with a fast CLI
Verified 2026-08-10
spec-kit
spec-driven agent workflows from GitHub Spec Kit
Verified 2026-08-10