MCP server for the Xahau network — offline Hook WASM VM + static analyzer, read-only ledger tools.
io.github.Hugegreencandle/xahau-mcp (MCP Server)
This Model Context Protocol (MCP) server targets the Xahau network and provides offline capabilities. It runs a Hook’s real WebAssembly bytecode in a local VM and performs Hooks-specific static analysis for security rules. It also includes read-only ledger access, Xahau-aware binary codec support, fee estimation, network-reward math, governance helpers, and unsigned-transaction builders.
🛠️ Key Features
Offline Hook execution via local VM for real WebAssembly bytecode
A Model Context Protocol server for the Xahau network with two firsts: it runs a Hook's real WebAssembly bytecode in a local VM (no xahaud node required), and it runs a Hooks-specific static-analysis / security rule engine over it — both fully offline. Around that core it adds read-only ledger access, a Xahau-aware binary codec, an instruction-count fee estimate, network-reward math, governance helpers, and unsigned-transaction builders.
Xahau is the XRPL fork whose flagship feature is Hooks — small on-ledger WebAssembly smart contracts. There was no MCP for Xahau and no static analyzer for Hooks; this is both.
xahau-mcp v2.0 flight simulator demo
The trifecta — safe Hooks, end to end
Three open-source tools, one workflow: write → simulate one tx → prove all inputs.
xahau-mcp is the simulate stage.
prove an invariant holds for every input in scope — or return the counterexample
Why it's useful
Point any MCP-capable agent (Claude, etc.) at this server and it can:
See the future before signing — simulate_transaction is a pre-sign flight simulator: every hook an unsigned transaction would trigger runs as real bytecode against live ledger state, with per-hook accept/rollback, decoded emitted transactions, simulated state writes and labeled static engine preflights. Its sibling what_if is a time machine: replay any real historical transaction — with your modifications — at its original ledger. Verified to reproduce a real claim's emitted GenesisMint payout to the drop (72,251,963 drops), test-locked.
Run a Hook without deploying it — execute_hook instantiates the real CreateCode WASM in a local VM, supplies the Hook API over a simulated transaction + ledger state, and reports the actual accept/rollback decision, return code/string, state writes, emitted transactions and a call trace. The first dev-accessible Hook simulator that needs no xahaud node.
Audit a Hook before it's installed — paste the CreateCode WASM (or an on-ledger hook hash) and get SARIF-lite findings: missing accept/rollback exit, unguarded loops (_g), unknown env imports, dangerous HookGrants, over-broad HookOn, and more.
Decode the cryptic HookOn bitmap in both directions — the 256-bit, inverted, active-low mask (with the active-high SetHook bit) is easy to get wrong; here it's verified and round-trip-tested.
Answer the #1 retail question — reward_status tells any account whether it's opted in to Xahau network rewards (Balance Adjustments), the exact XAH accrued — computed with the genesis reward hook's own formula and live parameters, verified to reproduce a real on-chain payout to the drop — when it can next claim, and whether the claim is overdue (late claiming forfeits yield).
Diagnose an Evernode host — evernode_host_diagnostics automates the official troubleshooting checklist for Xahau's largest operator group: registration, heartbeat liveness (the actual on-chain active rule), reputation, EVR trustline, lease offers, specs and accumulated rewards, in one read-only call.
Explain a failed transaction — diagnose_failed_tx turns an engine result + hook return strings into a plain-English cause and a concrete fix.
Watch governance live — governance_state decodes the Genesis Governance Game's full hook state: who holds the 20 seats, every open vote and tally, and whether a change (member swap, reward-rate change) is about to be actioned. No explorer shows this.
Build unsigned transactions (SetHook, ClaimReward, Payment) with an automatic security preflight — returned unsigned, to be signed offline.
Why this is the most advanced blockchain MCP we know of
Strong claim, so here is the checkable evidence (2026-06-11). To our knowledge no MCP for ANY
chain — Ethereum, Solana, Bitcoin, XRPL or otherwise — combines even two of these; the closest
comparators are cloud-simulation MCPs (e.g. Tenderly's, which simulates on their hosted
infrastructure) and standalone analyzers (e.g. Slither, which is EVM-only and not an MCP):
Executes real on-chain contract bytecode in a LOCAL VM — execute_hook runs the actual
CreateCode WASM with no node, no cloud, no account. Not an ABI wrapper, not a hosted simulator.
Publishes a measured, regression-locked fidelity score against chain ground truth —
vm_fidelity_report replays 30 real mainnet hook executions: 30/30 agree (100%), 0 degraded,
including the foreign-state-reading hook that dominates live traffic. Those 30 are all
accept-direction (live Xahau traffic is heartbeat-dominated), and the metric says so itself —
it reports the accept/rollback composition and warns that an accept-only corpus can't distinguish
the VM from an always-accept stub. The rollback direction is exercised on real genesis bytecode
(governance Invoke → rollback) in tests/regression.test.ts.
The corpus, the method and the honest history (25% → 0% → 100%) are in
docs/FIDELITY.md. We know of no other blockchain MCP that even attempts this.
In-protocol static security analysis — a Hooks-specific rule engine (SARIF-lite findings),
calibrated against the network's own genesis hooks.
In-protocol differential fuzzing — fuzz_hook maps a contract's accept/reject decision
boundary in the local VM.
Post-mortems real transactions with real bytecode — hook_execution_postmortem replays what
actually fired on chain and compares.
Reproduces on-chain economics exactly — reward_status re-implements the genesis reward
hook's formula and reproduces a real emitted payout to the drop (verified, test-locked).
Decodes live governance end-to-end — governance_state shows every seat, vote, tally and
threshold of the Governance Game, live.
Operational doctors for the ecosystem's real pain: failed-tx diagnosis with cause+fix,
Evernode host health, claim-overdue detection.
Every claim above is reproducible from this repo: the corpus is committed, the tests assert the
numbers, and the canonical sources (xahaud genesis hooks, evernode-js-client) are cited in code.
Safety posture
Read-only toward the network. There is no submit and no sign anywhere in this server.
No key custody. Builder tools never accept a secret/seed and always return an unsigned transaction plus instructions to sign offline (e.g. with xaman or xrpl-accountlib). They default to testnet.
Honest fidelity.execute_hook runs the real bytecode against a simulated environment. The VM implements a large slice of the 78-function Hook API — nearly the whole XFL float API (verified against float_one; float_root/float_log are the exceptions — they return NOT_IMPLEMENTED), the slot table + STObject subfield extraction (slot_subfield/sto_subfield, byte-exact against real txns), state, otxn_*/hook_*, util_accid/util_raddr/util_verify/util_sha512h, and more. STObject mutation (sto_emplace/erase/validate), util_keylet (account + hook verified against live ledger indexes; offer/escrow/check/ticket/signers canonical + fail-safe), slot_set + foreign hook state (state_foreign/state_foreign_set) with async pre-resolve (execute_hook resolveKeylets:true fetches the ledger objects AND foreign-state entries the hook reads — iteratively, since one resolved read can expose the next — and re-runs), slot_float/float_sto/float_sto_set (STAmount ⇄ XFL both ways — parse a native or issued/IOU Amount into a float and serialize one back; the issued layout below bit 63 is the XFL layout), and 32-byte state-key padding (short keys are left-zero-padded exactly as on-ledger) are now supported. state_foreign_set records the write but does NOT model the on-chain HookGrant requirement; etxn_details serves a disclosed SYNTHETIC placeholder (listed in syntheticCalls, cannot change the accept/rollback decision). What still can't be faithful is honestly recorded: unverified keylet subtypes, meta_slot, and other un-modelled calls return the real NOT_IMPLEMENTED code, are listed in unsupportedCalls, and mark the run degraded — never faked. The VM models the guard budget (_g enforces each guard's declared maxiter → GUARD_VIOLATION), and reports stateApplied (state writes commit only on accept, discarded on rollback). It is not a consensus-faithful xahaud replica — it has no fee/fuel metering beyond guards, XFL math truncates rather than round-half-up (so float_mulratio's round-up flag and last-significant-digit results can differ), value-level math is verified only where tested. Hooks with a loop but no _g guard are refused before execution (invalid on-chain), and guarded runs are bounded by a VM budget (1M cumulative guard calls / 2s wall clock — labeled as a local VM cap, not a consensus limit); always confirm financial/resource hooks on testnet. hook_dry_run is STATIC_ONLY, compute_reward is DOCUMENTED_MODEL (legacy — prefer reward_status, whose REWARD_HOOK_FORMULA re-implements reward.c exactly and reproduces a real on-chain GenesisMint payout to the drop), estimate_hook_fee is ESTIMATE.
Resources & prompts. Beyond tools, the server exposes MCP resources (xahau://rules, xahau://hook-api, xahau://tx-types) and guided prompts (audit_hook, simulate_hook, explain_hook) so agents can pull reference data and run the common workflows directly.
Tools
Hook intelligence (offline — the core)
Tool
Purpose
execute_hook
Run the real Hook bytecode in a local VM against a simulated tx/state → actual accept/rollback, return code, state writes, emits, trace (LOCAL_VM).
simulate_transaction
PRE-SIGN FLIGHT SIMULATOR — predict an unsigned tx's fate: originator + stakeholder hook chains (order canonical from xahaud Transactor.cpp/applyHook.cpp) run as real bytecode against live state; per-hook verdicts, decoded emits, state writes, static engine preflights, scam score.
what_if
TIME MACHINE — fetch a real historical tx, apply your overrides, re-simulate at its original ledger. Reproduces the real reward claim's GenesisMint to the drop (test-locked).
fuzz_hook
Differential fuzzing: sweep many generated transactions through the VM to map the hook's accept/rollback decision boundary (which tx types / amounts it accepts vs rejects).
annotate_hook_trace
Decode an execute_hooktrace[] into human-readable values by byte-width: canonical XFL float (definite), int64/native-drops (both endians), UInt32 + Ripple-epoch date, candidate account-id → r-address (possible), 32-byte hash. Raw hex always preserved; offline.
Post-mortem a real on-chain tx's hooks: fetch the tx + its meta.HookExecutions + engine result, then run each fired hook's real bytecode through the VM and compare the VM's accept/rollback to what the chain recorded. On-chain decision is authoritative; VM run is LOCAL_VM; agree is null (not false) when degraded/indeterminate. Serial RPC: 1 tx + 1 ledger_entry per unique HookHash.
vm_fidelity_report
Honest fidelity metric: replays a committed corpus of real mainnet HookExecutions through the VM and reports agreement % over comparable (non-degraded) runs only; offline.
classify_hook
Infer in plain English what a hook does (firewall/emitter/stateful/financial/…).
Engine result code ⇄ name (e.g. 153 ⇄ tecHOOK_REJECTED).
diagnose_failed_tx
"Why did my transaction fail?" — plain-English diagnosis from on-chain facts: engine result → cause + concrete fix (~30-code catalog), hook rollback return-strings decoded and interpreted (the reward hook's "You must wait N seconds" becomes a claimable-at date), the partial-payment trap on "successful" Payments, and not-found triage (expired LastLedgerSequence / wrong network). 1 read.
Decode a Xaman txjson or tx_blob → plain-English "what you authorize" + safety warnings.
decode_lease_uri
Decode an Evernode lease URIToken (evrlease/LTV) → lease index, EVR amount (XFL), ToS hash, IP.
evernode_host_diagnostics
One-call Evernode host health check (the official troubleshooting checklist, automated): registration entry, heartbeat liveness vs the on-chain active rule, instance load, reputation, EVR trustline/balance, registration URIToken, lease offers, machine specs + accumulated EVR reward. Layout verified against the canonical evernode-js-client + a live mainnet host (~9 serial reads).
Score a sign request (txjson or tx_blob) for risky patterns → dangerScore 0-100 + SAFE/CAUTION/DANGER tier + per-rule findings (SetHook, AccountDelete-to-other, regular-key/signer-list changes, large native payment, no-expiry, pre-signed). Offline heuristic on tx shape only — every finding is a potential risk, never a confirmed scam; no block-list lookup, no on-chain malice check.
Balance Adjustment doctor — opted in? exact accrued XAH (the genesis reward hook's own formula from reward.c, with live RR/RD read from genesis hook state; reproduces a real on-chain GenesisMint payout to the drop), next-claim countdown, overdue-claim warning (late claiming forfeits yield), plus an unsigned opt-in/claim ClaimReward when applicable (3 serial reads).
Grade an account for quantum (HNDL) readiness — master-key/regular-key/multisig + hooks → score, tier, recommendations (with a Hook/PQC angle).
governance_state
Full live decode of the Governance Game: all 20 seats + members, member count, live reward rate/delay, every open vote (who voted what) and every tally with its threshold (80% membership / 100% else) and reached-flag. Layout canonical from xahaud hook/genesis/govern.c.
decode_b2m
Burn2Mint classification.
Amendment intelligence (read-only, no admin node)
Tool
Purpose
get_amendment_status
Every enabled amendment (resolved to its human name where known, raw hash otherwise) + any in the voting set (already >80%, counting toward enablement). Reads the on-ledger Amendments singleton — no admin feature call needed.
predict_amendment_activation
For each amendment holding majority: when it reached >80% and its estimated enable time (majority + Xahau's 5-day window), with days remaining and an eligible-now flag.
check_amendment_blocked
Is the configured node amendment blocked (xahaud too old to read an enabled amendment)? Build version, validated ledger, upgrade remedy.
diff_node_amendments
Diff enabled-amendment sets across two networks (e.g. mainnet vs testnet) — explains why a tx/Hook feature works on one but not the other.
Audit / forensics
Tool
Purpose
trace_transaction_stakeholders
Every account a tx touched, from metadata. The Touch amendment forces all stakeholders into metadata, so this is the authoritative participant list (roles + entry types + materially-changed flag).
verify_double_threading
Structural audit of a tx's metadata threading; flags the duplicate-node symptom fixProvisionalDoubleThreading addressed.
audit_account_remarks
Decode the Remarks attached to an account's objects (name/value hex→text, immutable flag) — dynamic-NFT and annotation patterns.
Unsigned builders (no keys, testnet-default)
Tool
Purpose
build_sethook_unsigned
UNSIGNED SetHook with automatic analyze_hook preflight.
UNSIGNED Remit (XLS-55) — atomic multi-asset push: multiple currencies (native + issued) and/or URIToken transfers and/or an inline URIToken mint to one destination, all-or-nothing. Canonical AmountEntry/MintURIToken nesting; auto trustline/reserve/account-create semantics; optional third-party Inform. Payload preflight included.
build_set_remarks_unsigned
UNSIGNED SetRemarks (Remarks amendment) — attach/update/delete key-value remarks on a ledger object you own/issue. immutable (tfImmutable) support, delete-by-omission, 32-max + uniqueness + size guards. Powers dynamic NFTs.
build_clawback_unsigned
UNSIGNED Clawback — issuer revokes issued tokens from a holder (Amount.issuer = holder, the canonical gotcha). Opt-in + XAH guards.
build_deepfreeze_unsigned
UNSIGNED TrustSet freeze toggle — deep_freeze (blocks send+receive), freeze, and their clears, mapped to the correct tfSetDeepFreeze/tfSetFreeze flags.
prepare_transaction
Autofill Sequence/Fee/LastLedgerSequence/NetworkID from the live network → ready to sign offline (never signs).
Cron (scheduled Hook execution)
Tool
Purpose
build_cronset_unsigned
UNSIGNED CronSet — schedule a Hook's future self-invocations: StartTime (Ripple-epoch, 0=ASAP) or startInSeconds, RepeatCount (0–256), DelaySeconds, or cancel (tfCronUnset). Preflight catches recurring-without-interval and no-Hook cases.
list_cron_jobs
Read an account's Cron ledger objects with decoded StartTime (ISO), DelaySeconds, remaining RepeatCount, and an estimated next-fire time (raw object included).
monitor_cron_health
Flag Crons whose remaining repeats are near exhaustion (≤ threshold) so a recurring governance/game Hook doesn't silently stop.
Install
New here or non-technical? Start with the plain-English tutorial — what it does + cool things to just ask.
Install straight from GitHub — no npm-registry account needed; it builds on install:
bash
npm install -g github:Hugegreencandle/xahau-mcp
Or clone and build:
bash
git clone https://github.com/Hugegreencandle/xahau-mcp && cd xahau-mcp
npm install # the `prepare` script compiles dist/ automatically
npm run smoke # health check + a live mainnet read
npm test# 298 tests (offline)
Also published to GitHub Packages as @hugegreencandle/xahau-mcp. GitHub Packages requires auth even for public installs, so add to your .npmrc:
code
@hugegreencandle:registry=https://npm.pkg.github.com
//npm.pkg.github.com/:_authToken=YOUR_GITHUB_TOKEN # token with read:packages
then npm install -g @hugegreencandle/xahau-mcp. (The github: install above needs no auth and is simpler.)
Add to an MCP client (e.g. Claude Code / Desktop):
json
{"mcpServers":{"xahau":{"command":"xahau-mcp"}}}
HTTP shim (browsers, wallets, web tools)
The MCP server speaks stdio — browsers, the Xaman webview, and a public "simulate any hook" web tool can't. The HTTP shim (src/http.ts) exposes the same flight-simulator core over plain HTTP. Read-only; never signs or submits.
bash
PORT=8787 npm run http # or: PORT=8787 node dist/http.js
curl localhost:8787/health
# or deploy the container (non-root, healthchecked):
docker build --target http -t xahau-mcp-http .
docker run -p 8787:8787 -e MAX_INFLIGHT=2 -e XAHC_HOOK_MEM_MB=128 xahau-mcp-http
On platforms that build the last Dockerfile stage (Railway, etc.), use Dockerfile.http (set RAILWAY_DOCKERFILE_PATH=Dockerfile.http) — it's the http stage as the final stage.
Endpoints
Method · path
Body
Returns
POST /simulate
{ tx, network?, ledgerIndex?, candidateCode? }
full Simulation — per-hook accept/rollback, decoded emits, state writes, static preflights
POST /what-if
{ txHash, overrides?, network? }
Simulation of a real historical tx re-run (with your overrides) at its original ledger
SandboxResult — run a hook's bytecode in isolation (offline, no RPC)
POST /analyze
{ wasmHex, hookOn?, namespace?, … }
{ findings, summary } — the static rule engine (offline, instant, the cheap top-of-funnel)
GET /fidelity
—
the VM fidelity report + corpusHash + lastRun (echoes the accept-only coverageWarning verbatim)
GET /health
—
{ ok, inflight, endpoints }
Simulate a not-yet-deployed hook — pass candidateCode (the built wasm hex) on /simulate. It replaces tx.Account's on-ledger hook chain, so a freshly-compiled hook runs against the live ledger + full TSH chain BEFORE you SetHook it. (This is what xahc verify and a public simulate tool call.)
Config (env)
Var
Default
Purpose
PORT
8787
listen port
RL_MAX
20
requests / IP / minute
MAX_INFLIGHT
4
concurrent simulations
XAHC_HOOK_TIMEOUT_MS
3000
hard wall-clock per hook execution
XAHC_HOOK_MEM_MB
256
memory cap per hook execution
XAHC_SIM_SPACING_MS
sim default
inter-read RPC spacing — 0 for your own node, ~300 behind a shared public node
TRUST_PROXY
unset
set only when behind a reverse proxy that sets X-Forwarded-For
XAHAU_RPC_URLS
data/endpoints.json
comma-separated mainnet RPC failover list
Running it publicly
The shim serves untrusted, user-supplied wasm, so each hook execution runs in a throwaway worker_thread with the timeout + memory cap above — a hanging or allocation-bombing hook is terminated without blocking or crashing the main process (security-audited; an infinite loop / unguarded recursion returns 422, not downtime). For a public deploy:
Run behind a process supervisor (systemd / pm2 / container restart policy) for defense-in-depth.
Only set TRUST_PROXY when genuinely behind a proxy that sets X-Forwarded-For; otherwise the rate limiter keys on the socket address (setting it lets clients spoof the header and bypass the limit).
Per-IP rate limit + a global concurrency cap are on by default — tune via env.
Security
Designed defensively and reviewed (npm audit + a danger-surface pass):
Read-only & no key custody — no sign/submit anywhere; builder tools never accept a secret and only emit unsigned transactions to sign offline.
No code-exec surface — no eval/Function, no child_process/shell, no filesystem writes, no dynamic require. RPC fetch only ever hits the fixed endpoints in data/endpoints.json (or your XAHAU_RPC_URLS / XAHAU_TEST_RPC_URLS overrides for mainnet/testnet) — never a URL built from tool input, so no SSRF.
Untrusted Hook WASM is sandboxed — execute_hook/fuzz_hook run hook bytecode in Node's WebAssembly engine, which has no syscall/fs/network access; a hook can only call the in-memory JS Hook-API shims, with bounds-checked memory reads/writes.
Untrusted-bytecode hardening: before executing a hook the VM refuses modules with an unguarded loop (more loops than _g guard call-sites), with an opcode-scan that couldn't verify the loops, over 128 KiB of bytecode, or declaring more than 512 memory pages; guarded loops are bounded by a cumulative guard budget + wall-clock cap. So an attacker-supplied hook can't hang or OOM a run. Tool output is data, not instructions (treat it as such, as with any MCP).
Dependencies:npm audit reports only low-severity advisories transitively under xrpl-accountlib's signing libraries (elliptic/bip32/tiny-secp256k1) — code paths this server never calls (it uses only the binary codec).
How it works
No heavy deps. Three runtime deps: @modelcontextprotocol/sdk, zod, and xrpl-accountlib (used only for the Xahau-aware binary codec; its signing surface is never called). RPC is plain fetch; the WASM reader is hand-rolled and zero-dep; the VM uses Node's built-in WebAssembly engine to run the bytecode with a JS Hook API shim — no WASM toolchain or native deps.
Real data, regenerable.data/ is built from a live Xahau node's server_definitions and the canonical Hook API list (Xahau/hooks-rsc/extern.h) via npm run fetch:all. The 78-function Hook API catalog carries per-function hazard metadata that drives the analyzer.
HookOn semantics are verified against the Xahau docs: 256-bit, bit n = tx type n, inverted/active-low (set = does not fire), with bit 22 (SetHook) active-high.