|
flowgate — Guardrails for AI coding agents that actually hold
Your AI coding agent deployed before tests and approved before review, and a longer prompt didn't fix it. flowgate puts agents on a state machine they can't...
|
84 % |
|
LogBear — give your AI coding agent memory of what broke
Your coding agent can't see your dev server crash. LogBear records every run on your machine and tells Claude Code, Cursor or Codex the moment something brea...
|
84 % |
|
Enact — The Agent Firewall for Claude Code, Cursor, and Codex
An AI coding agent wiped a production database in 9 seconds while trying to fix an unrelated bug. PocketOS, April 2026. Same shape as Replit and DataTalks. E...
|
83 % |
|
AI Agent Observability & Monitoring - Latitude
Open-source AI agent monitoring platform. Full observability into what's failing in production. Discover underlying issues, get alerts when something breaks...
|
83 % |
|
AI Agent Sync — Keep your AI honest and on task.
A guard for your AI coding agent. Re-injects what it forgot after a compaction, and catches it claiming things it never checked. Installs in one command. Run...
|
83 % |
|
onpath — Know when your AI coding agent is stuck
onpath monitors AI coding agent sessions locally, detects repeated failures and stalled progress, and gives teams clear recovery context.
|
83 % |
|
Platypus AIlytics — Private AI that learns how your operation runs
Platypus AIlytics builds private agentic AI that learns your entire process and answers like your best expert — on hardware you own, inside your network or c...
|
83 % |
|
Runheed: Error tracking your coding agent can read
Runheed sends your app's production errors to Claude Code, Cursor or any MCP client, so your agent can read the stack trace, fix it and resolve it. From $29/mo.
|
83 % |
|
snapfail — Browser error context for your AI editor
snapfail captures stack traces, console logs, network requests and user replay when your app crashes — and packages it as structured evidence your AI agent r...
|
82 % |
|
Lamina: Transactional product graph for AI agents
A transactional product graph for AI coding agents. Design actors, states, flows, and failures, then verify the live product with persona missions.
|
82 % |
|
Sailfish AI
|
82 % |
|
Instify — Run AI coding agents like a product team
Stop losing features in an AI chat. Instify puts Claude Code and Codex on a real board — plan as stories, run an agent, review every change. Mac-native, free.
|
82 % |
|
QualityGuard.ai | Autonomous QA agent for E2E and drift detection
QualityGuard.ai runs autonomous E2E, log, and metric checks to detect product drift, open tasks with evidence, and keep quality high while teams ship fast.
|
82 % |
|
Dunetrace | Runtime reliability for AI agents
Dunetrace catches AI agent failures at runtime, alerts on Slack, finds the root cause, and fixes them. Structural and semantic detection, runtime prevention,...
|
82 % |
|
Fugue Labs: agent runtimes in Go
An AI research lab building production agent runtimes. Gollem: the Go framework with runtime traces you can resume, fork, diff, and use as evidence. Sleepy:...
|
82 % |
|
sentrux — The sensor that helps AI agents close the feedback loop
Recursive self-improvement of code quality. Real-time architectural sensor for AI coding agents. 52 languages. One score. Pure Rust.
|
82 % |
|
Devvy — Engineering that runs itself
Turn messages into production code. Devvy spins up isolated Firecracker VMs from Slack, writes the code, and opens a pull request.
|
82 % |
|
AgentDbg | Local-first debugger for AI agents
Debug AI agent runs locally with a step-by-step timeline of LLM calls, tool calls, errors, and loop warnings. Stop runaway runs with built-in guardrails.
|
82 % |
|
Agentsure — AI Incident Response & Recovery, retained.
A 24/7 incident response and recovery retainer for engineering teams running AI agents in production. Engineers in your war room, not a vendor portal.
|
82 % |
|
Unravel - debugger-first observability for PHP applications
A webhook, the jobs it fired, the broadcast and the exception, captured as one causal trace. Connect your AI agent over MCP and it fixes the bug with the who...
|
82 % |
Detected social media networks