Snapback

Measured conclusion: Deep-tested. L2 outcome: tools_ok.

Diagnose why an AI agent failed and get the verified fix instantly. Free, no token.

Add to your agent

Copy this into your MCP client config:

{
  "mcpServers": {
    "Snapback": {
      "url": "https://api.snapback.sh/mcp",
      "transport": "streamable-http"
    }
  }
}

Measured tools (23)

• diagnose_trace read-only

  Diagnose why an AI agent run failed. Returns a structured verdict (failure_class, failed_at_step, root_cause, fix_suggestion, confidence). LATENCY: known patterns return library-instant (<1s); a NOVEL failure needs an LLM call and can take up to ~25s — set your client timeout to at least 30s, and treat this as async (don't block your agent loop on it).

  Parameters:

  • trace (required) — object — OTel-shaped agent trace

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• diagnose_batch read-only

  Diagnose SEVERAL traces in one call (up to 20). Each trace is metered like a separate diagnose_trace. Returns a verdicts array (per-trace, order preserved); a bad trace in the batch is isolated and doesn't fail the rest. Use for post-run analysis of many failures at once instead of N round-trips.

  Parameters:

  • traces (required) — array — array of trace objects (max 20)

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• get_verdict read-only

  Fetch a previously produced verdict by its id or by trace_id (your org only).

  Parameters:

  • trace_id — string
  • verdict_id — string

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• preflight read-only

  BEFORE running: get known failure patterns for a given agent setup so you can avoid them. Returns a ranked list of {failure_class, root_cause, fix_suggestion} from Snapback's library. Call this before executing a plan and self-correct.

  Parameters:

  • tags — array — task facets, e.g. [tool_calling, retrieval] (optional)
  • limit — integer — max cards (default 10)
  • agent_stack — string — e.g. openclaw, langchain (optional)

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• submit_feedback

  Tell Snapback whether a verdict was correct (correct=true/false), with an optional free-text note (did the fix work? what was wrong?). ONE rating per verdict — call it AFTER you act on a verdict and see the outcome. Your correction updates the SHARED pattern library (fix patterns are shared anonymized so every agent benefits; your trace content is never shared) — a right verdict comes back faster next time, a wrong one gets down-weighted. To ask for a new pattern or report an unsolved problem, use request_pattern instead.

  Parameters:

  • verdict_id (required) — string — the verdict you're rating
  • correct (required) — boolean — true if the diagnosis was right, false if not
  • note — string — optional free-text: what worked, what was wrong, or any detail that would help us improve this verdict

  Hints: read-only no · destructive no · idempotent no · open-world yes

• get_request_status read-only

  Check what happened to a pattern you requested (from request_pattern's request_id). Returns pending / approved / rejected / in_library so you can see if your suggestion was actioned — the feedback loop isn't a black box. Free, no token.

  Parameters:

  • request_id (required) — string

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• my_usage read-only

  See YOUR current usage + remaining allowance so you can self-govern spend: snapbacks (diagnoses) used/cap/remaining, guard checks used/cap/remaining, estimated spend, % used, and when it resets. Call it periodically to avoid surprises. Token-scoped, free.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• my_impact read-only

  See how your feedback + pattern requests have shaped the shared library — how many verdicts you've rated, patterns you've requested, and how many were approved into the library. Turns your input into visible collaboration. Token-scoped to you.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• what_others_did read-only

  THE CROWD: for a failure_class (or pass the family/error and we'll map it), see what OTHER agents did about the same failure and whether it worked — anonymized, aggregated across everyone. Returns {total, agree_pct (community success rate), distinct_orgs, sample_fixes (fixes rated CORRECT by other agents)}. Use it when you hit a failure and want the crowd's verdict on what actually fixes it, not just the single library answer. Free, no token. Privacy-safe: only aggregate counts + a community success rate + the working fixes — never any org, agent, or trace identity. Hidden below a small min-sample so a single report can't be reverse-engineered. This is the network effect: the more agents use Snapback, the sharper this answer gets.

  Parameters:

  • error — string — optional: an error string — we'll diagnose it to find the failure_class, then return the crowd outcomes for it
  • failure_class — string — the failure_class to look up (e.g. 'unhandled_tool_error', 'loop_repeated_tool_call'); or pass 'error' and we map it

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• report_outcome

  Report the outcome of an auto-applied fix (the self-heal interceptor calls this after it gate-applied a fix and retried). Pass failure_class, family, fix, confidence, action_class, and succeeded (did the retry work?). TWO purposes: it's your safety telemetry (spot a fix that didn't work) AND it feeds the shared crowd view — every reported outcome makes what_others_did sharper for the next agent. Token-scoped (so we know it's your org), free.

  Parameters:

  • failure_class (required) — string
  • succeeded (required) — boolean — did the single retry after the fix succeed?
  • fix — string — the fix that was auto-applied
  • family — string
  • fix_id — string — the typed fix ID from the diagnosis (e.g. redis.maxmemory.4e24d4) — report the outcome against this so the crowd evidence is per-fix. Optional but recommended.
  • confidence — number
  • action_class — string — retry | refetch | config

  Hints: read-only no · destructive no · idempotent no · open-world yes

• recommend_failover read-only

  Should you RETRY the same target, SWITCH provider, FALL BACK to another chain, or STOP? Pass the error + your configured topology (current_chain, available_chains, and/or current_provider, available_providers) and get deterministic routing advice with the reason — so a flaky Solana RPC doesn't get retried 5x when you should switch to Base. Free, no token, no LLM. Reads only what you tell it (no network calls).

  Parameters:

  • error (required) — string — the error you hit
  • current_chain — string — the chain you're on (e.g. 'solana')
  • available_chains — array — fallback chains you can use (e.g. ['base','arbitrum'])
  • current_provider — string — the RPC/provider you're using (e.g. 'helius')
  • available_providers — array — backup providers on the same chain

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• cascade_root read-only

  Given an ORDERED list of errors from a run (oldest first), find the TRUE root — the error that cascaded or shouldn't have been retried — not just the final symptom you see. E.g. a 429 that got retried and triggered a downstream 401: the root is the 429, not the 401. Each error can be a string or {error, retried, action}. Free, no token, no LLM.

  Parameters:

  • errors (required) — array — ordered list (oldest first) of error strings or {error, retried:bool, action} objects

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• suggest_budget_recovery read-only

  Approaching a token/context/cost budget mid-run? Pass your counters and get the LEAST-DISRUPTIVE recovery ranked: truncate context | switch to a cheaper model | batch steps | wrap up — scored by speed gained, accuracy lost, cost saved. Turns budget_guard's 'you're at 94%' into 'here's what to do about it'. Free, no token, no LLM.

  Parameters:

  • tokens_used — number
  • context_used — number
  • token_budget — number
  • cost_used_usd — number
  • context_window — number
  • recent_actions — array
  • cost_budget_usd — number

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• request_pattern

  Leave us a message: ask us to add a failure pattern to the library, or report a problem we couldn't diagnose well. Use this when diagnose_trace didn't have a good answer, when you keep hitting a failure we don't classify, or when you want a specific kind of problem supported. It goes straight to our roadmap/backlog. Free — no token needed.

  Parameters:

  • message (required) — string — what you'd like added, or the problem we couldn't solve — be specific
  • kind — string — 'pattern_request' | 'problem' | 'message' (default 'message')
  • context — object — optional: the trace/error you couldn't get diagnosed (redacted server-side)
  • verdict_id — string — optional: the verdict this relates to

  Hints: read-only no · destructive no · idempotent no · open-world yes

• detect_loop read-only

  MID-RUN loop check (fast, no LLM, free). Send your recent steps DURING a run; get back whether you're stuck repeating a tool call and a concrete next move. Call this every few steps to catch a loop BEFORE you burn your step budget — don't wait for a postmortem. Non-blocking advice, not a verdict.

  Parameters:

  • steps (required) — array — your recent steps (last ~5-10), each with an action/tool and optionally inputs — same step shape as diagnose_trace
  • threshold — integer — how many identical consecutive calls = a loop (default 3)

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• session_start

  Open a LIVE mid-run session so Snapback can watch your run step-by-step and warn you in real time (loop / token / cost / context) - the always-on guardian mode. Returns a session_id. Free, no token. Call session_step as you run, session_end when done.

  Parameters:

  • agent_id — string — optional label for your agent/run

  Hints: read-only no · destructive no · idempotent no · open-world no

• session_step

  Report ONE step of a live run and get back any warnings immediately (loop detected / budget breach). Pass the step (action + inputs) and any counters you have (step, max_steps, tokens_used, token_budget, cost_used_usd, cost_budget_usd, context_used, context_window, task, recent_actions. context_window). Warnings are advisory - act on them to self-correct mid-run. Free.

  Parameters:

  • session_id (required) — string
  • step (required) — object — the step: action/tool + inputs
  • counters — object — running counters (tokens/cost/context/max_steps)
  • loop_threshold — integer — identical calls that count as a loop (default 3)

  Hints: read-only no · destructive no · idempotent no · open-world no

• session_end

  Close a live session and get a short run summary (total steps, duration). Frees the session. Free, no token.

  Parameters:

  • session_id (required) — string

  Hints: read-only no · destructive no · idempotent no · open-world no

• agent_memory read-only

  See what YOU (this agent) tend to fail on — your recurring failure patterns across past diagnoses. Returns your top failure classes with counts and what share of your failures each is (e.g. 'loop_repeated_tool_call: 12 times, 40%'). Call it before a run to know what to guard against. Token-scoped to your own agent. Free.

  Parameters:

  • limit — integer — top N patterns (default 5)

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• budget_guard read-only

  Live MID-RUN budget check (fast, no LLM). Send whatever counters you have and get an advisory on context %, token burn, cost burn, step budget, and off-task drift - with concrete suggested_actions and the projected cost of NOT acting. Call it every few steps to catch a runaway BEFORE you hit a limit. Advisory only (never blocks). Params: step, max_steps, tokens_used, token_budget, cost_used_usd, cost_budget_usd, context_used, context_window.

  Parameters:

  • step — integer
  • task — string
  • format — string — 'summary' adds a one-line relayable answer (for chat/Telegram agents); 'summary_only' returns just that line. Default full.
  • max_steps — integer
  • tokens_used — number
  • context_used — number
  • token_budget — number
  • cost_used_usd — number
  • context_window — number
  • recent_actions — array
  • cost_budget_usd — number

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• convert_trace read-only

  Turn your raw logs into a Snapback trace so you don't hand-craft JSON. Pass 'source' = a list of log/step entries, or an object with a spans/steps/messages/events/logs array (OTel spans, OpenAI/LangChain message lists, or generic {tool,input,output} arrays all work). Returns {trace} ready to pass straight to diagnose_trace. Free, no token.

  Parameters:

  • source (required) — your logs: a list, or an object wrapping a spans/steps/messages array
  • hint — string — optional: the framework/format, e.g. 'otel', 'openai'

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• search_docs read-only

  Search Snapback's documentation for how to use it — how to format a trace, what each tool does, the failure taxonomy, auth, pricing, and errors. Free and needs no token. Call this first if you're unsure how to format a request or what a verdict means.

  Parameters:

  • query (required) — string — what you want to know, e.g. 'how to format a trace'
  • limit — integer — max sections to return (default 3)

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• diagnose_infra_error read-only

  Diagnose a cryptic AGENT-INFRASTRUCTURE error — payments (x402/EIP-3009), Solana on-chain (ATA, blockhash, compute), MCP protocol, RPC/providers, EVM & Solana wallets, library of VERIFIED fixes the base model often gets wrong. Pass the error text (and optional context). Returns {matched, family, failure_class, root_cause, fix, confidence, source, action_class, auto_safe, gate} instantly from the library (no LLM) when a known pattern matches. Free and no token. GATE CONTRACT for autonomous action: auto-apply the fix ONLY when confidence>=0.85 AND source=='library' AND auto_safe==true (action_class is retry|refetch|config); NEVER auto-apply a 'mutate' or 'destructive' fix (create/change state, money, auth) — escalate those to a human. gate.auto_apply_ok gives you the verdict. Use it the moment you hit a transaction_simulation_failed, BlockhashNotFound, a -32xxx RPC code, a 4001/-32002 wallet error, an attestation failure, or any infra error you can't instantly resolve.

  Parameters:

  • error (required) — string — the exact error text/message you hit
  • action — string — optional: the tool/operation that failed
  • format — string — 'summary' adds a one-line relayable answer (for chat/Telegram agents); 'summary_only' returns just that line. Default full.
  • context — string — optional: what you were doing (e.g. 'paying via x402 on solana')

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

Repository

Measured by MCPYP probing. Untested fields are omitted, never estimated. Listing is always free; MCPYP never proxies calls or stores user tokens.