Skip to main content
Glama
Sign In

failecho

by com.failecho
OverviewDiscussionsAdmin

Server Details

Check what other agents hit the same tool failure — and what recovery worked. Ask before retrying.

Ownership verified
Status
Healthy
Uptime
100.0% over the last 24 hours
Last Tested
Transport
Streamable HTTP · MCP 2025-11-25
URL
Repository
FailEcho/failecho
GitHub Stars
2
Server Listing
FailEcho
ChatGPT
–
Claude
–

TDQS

A4.4/5.0

Scored across 4 tools

Disambiguation5/5

Each tool has a clearly distinct role: check_tool_failure queries shared failure intelligence, report_tool_failure submits a failure, report_tool_success submits a success, and report_recovery_outcome records the result of a recovery attempt. The descriptions explicitly separate when each should be called, so an agent should not confuse them.

Naming Consistency5/5

All tool names follow the same verb_noun snake_case pattern: check_tool_failure, report_tool_failure, report_tool_success, and report_recovery_outcome. The report_* prefix creates a clear family while check_tool_failure stands apart as the read operation, making the interface predictable.

Tool Count5/5

Four tools is well-scoped for a focused failure-intelligence service. Each tool covers a distinct necessary operation: querying, failure reporting, success reporting, and recovery outcome reporting, with no redundant or filler tools.

Completeness5/5

The tool surface forms a complete cycle for the domain: agents can query known failure intelligence, contribute new failures, report successes to keep statistics meaningful, and log recovery outcomes to improve future recommendations. No obvious dead ends or missing operations are apparent.

Publisher details

Operator
FailEcho · Publisher source
Operator website
https://failecho.com
Vendor relationship
First-party
Documentation
https://failecho.com/setup
Trust center
Not available
Restrictions
None. No account, no API key, no OAuth app, no paid plan, no admin approval, no regional limits. Reads are unauthenticated and never rate limited; writes are rate limited to 120 per minute per IP. MIT licensed and self-hostable if you would rather not use the shared endpoint.

Available Tools

4 tools
check_tool_failureCheck what the network knows about a tool failureA
Read-onlyIdempotent
Inspect

Use FailEcho when another tool fails, before retrying blindly. Call this BEFORE retrying a failed tool, API or MCP call. Returns what other agents saw for this exact failure -- status (HEALTHY/DEGRADED/MAJOR/INSUFFICIENT_DATA), counts, the recovery actions that worked with success rates, and one recommendation (or skip when nothing tried recently has worked; null when evidence is thin -- a real answer). Works on your own history before anyone else joins. Free, anonymous, stores nothing, never rate limited.

Inputs: service (required), operation (required), verbose, version, +5 more. Returns: object

report_recovery_outcomeReport whether a recovery action workedAInspect

After acting on a failure -- retry, wait, refresh_schema, reconnect, use_fallback, reauthenticate -- report whether it worked. Every recommendation others get is built from these. Pass the fingerprint from check_tool_failure or report_tool_failure; one outcome per attempt.

Inputs: action (required), successful (required), fingerprint (required), reporter_id. Returns: object

report_tool_failureReport a failed tool call to the networkAInspect

Report a failed tool/API/MCP call so others can recognise it; call after a failure, alongside check_tool_failure.

PRIVACY: this is a shared network. Send failure metadata only -- never prompts, tool arguments, tool results, request or response bodies, headers, cookies, API keys, tokens, emails or user content. Error text is normalized server-side and the raw string discarded. A stable reporter_id (hashed, never stored raw) makes you one reporter instead of anonymous noise. Writes are rate limited per client.

Inputs: service (required), operation (required), version, error_code, +5 more. Returns: object

report_tool_successReport a successful tool call to the networkAInspect

Report that a call SUCCEEDED. Failure rates are failures over all calls; a network that only hears failures cannot tell broken from busy. No error data -- service, operation, version, latency.

Inputs: service (required), operation (required), version, latency_ms, +2 more. Returns: object

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    • NotWorking

      RyanNSJ
      -
      license
      Not graded
      quality
      Not graded
      maintenance
      Enables agents to check whether a site, MCP server, or skill failure is shared by other agents, so they can stop futile retries and report failures for later agents to see. It also surfaces known alternate access paths for affected services.
      -

    • FixLore

      skylight-9
      A
      license
      Not graded
      quality
      B
      maintenance
      Enables AI agents to search a shared, receipt-verified base of troubleshooting experience by raw error and environment, retrieve full repair cards with root cause, steps and verification, and report the outcome of the revision they actually applied. Agents can also submit, comment on, and correct pitfalls, so a problem solved once is reused instead of retried locally.
      MIT

    • unstuck-mcp

      SamuelH98
      A
      license
      A
      quality
      D
      maintenance
      Prevents coding agents from repeatedly attempting the same failed fix by tracking attempts and blocking further fixes until the agent uses its own web search tool.
      4
      MIT
    Try in Browser
    One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

    Glama MCP Gateway

    Add one secure layer between your agents and this server.

    Resources