My MCP server used to simply stop working. No crash dump. No stack trace in the logs. Clients connected without complain

My MCP server used to simply stop working. No crash dump. No stack trace in the logs. Clients connected without complaint, then after a few hours the whole thing went mute. Requests disappeared and the AI agent on the other end received nothing but blank air.

This is a frustratingly common story in the Model Context Protocol (MCP) ecosystem. The protocol defines how AI agents discover and call external tools, but the specification assumes you will handle errors yourself. Most tutorials and starter implementations skip past that part. They focus on the happy path: annotate a function, expose it through the server, and return a clean result. They rarely show you what happens when a network blip hits your external API, or when the model hallucinates a parameter name and sends garbage input. The result is a brittle server that looks healthy but has actually been dead for hours.

Why Blank Responses Are Worse Than Crashes

When an unhandled exception slips through in an MCP tool handler, the transport layer often swallows it. The server process stays alive, the socket remains open, but the client gets an empty response. This is more dangerous than a loud crash because your monitoring might not notice. The process is still running. The port is still listening. Yet every tool call returns nothing.

The AI model does not interpret silence as failure. It interprets silence as a successful call that produced no data. That blank response trains the model to improvise. It starts hallucinating facts to fill the gap, or it enters a loop of retrying the same broken call. Small issues like a transient network timeout or an invalid tool argument should never be allowed to cause this kind of behavior.

The Wrapper Pattern: Three Lines of Defense

I fixed this by wrapping every tool handler in a thin error-recovery layer. The wrapper does not try to predict every possible failure. It categorizes them and responds accordingly.

ConnectionError and TimeoutError
These arise when your server talks to an external API and the network wobbles. The instinctive fix is to restart the entire MCP server process. Do not do that. Rebooting drops active client connections, clears any in-memory state, and forces a full re-initialization. Instead, catch the connection failure and reconnect only the transport layer or HTTP client your tool uses. The server stays warm and ready for the next request immediately.

ValueError
This is what you see when the AI client sends malformed arguments. Maybe the model invented a parameter, passed a string where an integer was required, or forgot a required field. If you let this bubble up unhandled, the client gets either a crash or a blank reply. Catch it inside the wrapper, then construct a clear, specific message that tells the model exactly what went wrong. Explain which parameter failed and what was expected. Most modern AI models will read that message and self-correct on the very next turn. A vague error wastes a reasoning cycle. A precise error fixes the problem immediately.

General Exceptions
Keep a safety net. If an error falls outside the categories above, log the details for yourself and return a clean, generic failure response to the client. This prevents one weird edge case from killing the session for everyone. The server survives, the client gets a signal that something failed, and you keep enough context in your logs to debug later.

The isError Flag Is Non-Negotiable

Here is the detail that actually determines whether your fix works. MCP responses include an isError boolean field. If an exception occurs and you return an error message without setting isError to true, the client treats that error text as a successful tool result.

Imagine your external API hits a rate limit. You catch the exception and return the string "API rate limit exceeded" but leave isError as false. The client passes that string into the model's context window as if it were real tool output. The model then tries to reason over that text as if it were data. It might quote the error in a summary, or worse, it might hallucinate relationships between that error text and other facts. You have turned a temporary infrastructure hiccup into a source of misinformation.

Daima weka isError kuwa true unaporudisha payload ya kosa. Hii inampa mteja ishara ya wazi kwamba wito wa zana (tool call) umefeli, jambo linalomruhusu modeli kuamua ikiwa ijaribu tena, iombe ufafanuzi, au ijaribu zana nyingine kabisa.

Jua Nini cha Kukamata na Nini cha Kusitisha

Usifunike seva yako nzima kwa try-catch isiyo na mwelekeo inayomeza kila kitu. Baadhi ya makosa yanamaanisha kuwa seva inapaswa kusimama mara moja. Ikiwa variable ya mazingira (environment variable) inayohitajika haipo wakati wa kuanza, au faili lako la usanidi (configuration file) limeharibika, hakuna kiwango chochote cha kukamata makosa katika kiwango cha ombi (request-level) kitakachosaidia. Tengeneza darasa maalum la ubaguzi (exception class) kwa makosa makubwa kama haya na uruhusu yasitishe mchakato (process).

Kanuni ni rahisi. Ikiwa kosa ni la muda mfupi au limejitenga na ombi moja tu, likamate na urekebishe. Ikiwa kosa linamaanisha kuwa kila ombi linalofuata limehakikishwa kufeli, acha seva ife kwa kelele. Kufeli haraka wakati wa kuanza ni bora zaidi kuliko seva inayofanya kazi kwa shida kwa siku nyingi katika hali ya kuharibika.

Ongeza Uwezo wa Uangalizi (Observability) Kabla Hujauhitaji

Mara tu unapokuwa na wrapper imewekwa, iunganishe na uandishi wa kumbukumbu uliopangwa (structured logging). Rekodi kila wito wa zana na matokeo yake katika mfumo wa JSON. Jumuisha jina la zana, hoja ghafi (raw arguments), ucheleweshaji (latency), na ikiwa ilifanikiwa, ikafeli, au ikajaribiwa tena.

Nidhamu hii inalipa haraka. Unapogundua ongezeko la makosa, unaweza kuchuja kwa zana na kubaini mifumo (patterns) ndani ya dakika chache. Labda API fulani ya nje inaanza kutoa muda mrefu wa kusubiri (timeouts) wakati uleule kila siku, ikielekeza kwenye dirisha la matengenezo yaliyopangwa ambalo hukujulia. Labda zana moja inapokea hoja zisizo sahihi (malformed arguments) mara kwa mara, ikionyesha kasoro ya uhandisi wa prompt (prompt engineering) upande wa juu. Kumbukumbu za maandishi ya kawaida (plain text logs) zilizofichwa kwenye stack traces hufanya kazi hii ya upelelezi kuwa ngumu. JSON iliyopangwa inafanya iwe rahisi sana.

Matokeo ya Uzalishaji (Production)

Nimetumia mfumo huu wa wrapper kwenye seva mbili za MCP za uzalishaji (production) kwa wiki tatu zilizopita. Katika kipindi hicho, nimeona makosa ya kimya (silent failures) sifuri. Kabla ya kuongeza wrapper, nilikuwa nikipata wastani wa kosa moja lisiloeleweka kila siku. Mfumo huu si mrefu, lakini athari yake ni kubwa kwa sababu unatenganisha kelele zinazoweza kudhibitiwa na matatizo ya kweli.

Makosa ya kimya yana gharama kubwa kuliko hitilafu (crashes). Hitilafu huchochea mfumo wako wa kutoa taarifa (alerting system). Kimya kinapunguza tu uaminifu. Siku moja wakala wako wa AI anarudisha data muhimu ya zana, na siku inayofuata anaanza kutunga mambo kwa sababu seva ilisimama kujibu saa nyingi zilizopita. Mfumo wa wrapper unafungia pengo hilo. Unauwezesha seva yako kuendelea kufanya kazi katikati ya misukosuko midogo, unaupa modeli muktadha wa kutosha ili kurekebisha makosa yake yenyewe, na unahakikisha kwamba jambo la hatari kweli linapotokea, unasikia mara moja.

Ikiwa unajenga zana za MCP leo, anza na wrapper na alama ya isError. Kila kitu kingine ni usafishaji tu.