FailEcho

Failecho — AI skill for Claude Code

AI community

When a tool or model call fails, your agent learns what fixed it for other agents, or that nothing has.

How to install Failecho

This entry records only its repository, not the path inside it, so there is no exact command to give. Open FailEcho/failecho and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Failecho does

When a tool or model call fails, your agent learns what fixed it for other agents, or that nothing has. Plugins for Claude Code and OpenCode, an MCP proxy, a Python wrapper. Open source, metadata only.

Alternatives in AI

Multica — Make humans and AI agents work as one team — open-source and self-hostable 47.8k ★
AIHawk — Open-source AI browser agent: describe any task in plain language and it autonomously browses, clicks, types a 30.3k ★
WeKnora — Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, an 26.9k ★

README

**Failure intelligence for AI agents and autonomous software.**
Before you retry, check the echo.

[Website](https://failecho.com) · [Connect an agent](https://failecho.com/setup) · [API reference](https://failecho.com/docs) · [Live network](https://failecho.com/network) · [llms.txt](https://failecho.com/llms.txt)

![Python 3.11+](https://img.shields.io/badge/python-3.11%2B-1e6b45) ![FastAPI](https://img.shields.io/badge/FastAPI-async-17211d) ![MCP](https://img.shields.io/badge/MCP-streamable%20http-c00010) ![License MIT](https://img.shields.io/badge/license-MIT-4a574f)

FailEcho is a cross-agent failure intelligence network. When a tool or model call fails, it tells your agent what fixed that exact failure for other agents -- or that nothing has, so it stops retrying. Agents share the shape of their failures and what fixed them (metadata only, never prompts or data); the next agent to hit the same failure gets the answer. Open source, no account.

An agent's Groq call fails with 429; FailEcho answers: try switch_model, worked 306 of 336. Two lab agents on the same model: the one without FailEcho retries and fails, the one with it switches model and answers correctly. Then a skip verdict on an exhausted quota, an honest no-clear-fix answer, the metadata-only payload, and the lab scoreboard with the rows where FailEcho does not help.

Every line is real output: live queries to the lab network, one twin pair replayed from the lab ledger (24 Sep), the wrapper's actual payload, and the lab scoreboard with p-values. Our own agents; independent reporters so far: 0.