io.github.jasonxkensei/xproof

Est-ce que io.github.jasonxkensei/xproof est sûr ?

Au dernier audit MCP Hub, le serveur MCP « io.github.jasonxkensei/xproof » est classé « Non encore testé ». Référencé dans l'annuaire. Seul le pentest comportemental fait foi, et il n'a pas encore été exécuté sur ce serveur. Le verdict est réévalué en continu ; consultez la fiche pour l'état à jour.

Méthode : MCP Hub analyse la description de chaque outil (couche 1) et, pour le verdict fort, exécute le serveur dans une micro-VM isolée avec leurres et réseau bloqué (couche 2). On publie des noms et des compteurs, jamais le contenu brut renvoyé par un outil.

Les 13 outils

• audit_agent_session≈ 666 tokens

  Certify an AI agent's work session on the MultiversX blockchain before a critical action (trade, deploy, data access). Returns a proof_id that an operator can use in its own compliance policy. This tool records certification; whether a missing or failed certification blocks execution is an operator-implemented policy decision. Governance: set reversibility_class='irreversible' for actions that cannot be undone — a policy violation is flagged if confidence_level < 0.95 on irreversible decisions. Schema: https://provebeforeact.com/.well-known/agent-audit-schema.json

  Paramètres, retour & exemple (11)

  Paramètres

  • context · object · optionnel
    Optional additional context (model version, environment, tool chain, etc.)
  • agent_id · string · requis
    Identifier of the agent making the decision
  • decision · string · requis · approved | rejected | deferred
    Agent's decision about whether to proceed
  • timestamp · string · requis
    ISO 8601 timestamp of when the decision was made
  • risk_level · string · requis · low | medium | high | critical
    Assessed risk level of the action
  • session_id · string · requis
    Unique session identifier (UUID or timestamp-based)
  • action_type · string · requis · trade_execution | code_deploy | data_access | content_generation | api_call | other
    Category of the action being certified
  • inputs_hash · string · requis
    SHA-256 of all inputs analyzed before making the decision
  • risk_summary · string · optionnel
    Optional brief risk analysis justifying the decision
  • action_description · string · requis
    Human-readable description of the specific action
  • reversibility_class · string · optionnel · reversible | costly | irreversible
    Governance: how reversible is this action? 'reversible' = can be undone at low cost, 'costly' = reversible but expensive (fees, slippage, delay), 'irreversible' = cannot be undone (on-chain settlement, data deletion, email sent). When 'irreversible', a confidence_level >= 0.95 is required to be policy-compliant.

  Exemple d'appel

  {"name": "audit_agent_session", "arguments": {"agent_id": "texte", "decision": "approved", "timestamp": "texte", "risk_level": "low", "session_id": "texte", "action_type": "trade_execution"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• certify_file≈ 237 tokens

  Create a blockchain certification for a file. Records the SHA-256 hash on MultiversX blockchain as immutable proof of existence and ownership. Cost: $0.01 per certification, paid in EGLD.

  Paramètres, retour & exemple (4)

  Paramètres

  • filename · string · requis
    Original filename with extension (max 255 chars)
  • file_hash · string · requis
    SHA-256 hash of the file (64 hex characters)
  • author_name · string · optionnel
    Name of the certifier (default: AI Agent, max 128 chars)
  • webhook_url · string · optionnel
    Optional HTTPS URL for on-chain confirmation callback

  Exemple d'appel

  {"name": "certify_file", "arguments": {"filename": "texte", "file_hash": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• certify_with_confidence≈ 608 tokens

  Create a staged blockchain certification with a confidence score. Use this when your decision builds progressively — certify at 60% (initial assessment), 80% (pre-commitment), and 100% (final decision). Each stage shares the same decision_id, creating an on-chain audit trail of the decision process. Governance: set reversibility_class='irreversible' for actions that cannot be undone — xproof will flag a policy violation if confidence_level < 0.95. Cost: $0.01 per certification.

  Paramètres, retour & exemple (9)

  Paramètres

  • who · string · optionnel
    Agent identity (wallet address, name, or agent ID)
  • why · string · optionnel
    Reason or instruction hash driving this decision
  • filename · string · requis
    Original filename with extension (e.g. decision.json, max 255 chars)
  • file_hash · string · requis
    SHA-256 hash of the decision or output file (64 hex characters)
  • author_name · string · optionnel
    Name of the certifying agent (default: AI Agent, max 128 chars)
  • decision_id · string · requis
    Shared UUID linking all confidence stages for the same decision. Generate once and reuse across all stages.
  • threshold_stage · string · requis · initial | partial | pre-commitment | final
    Named stage of the decision: initial (first assessment), partial (gathering info), pre-commitment (almost certain), final (committed).
  • confidence_level · number · requis
    Confidence score from 0.0 to 1.0. Typical values: 0.6 (initial), 0.8 (pre-commitment), 1.0 (final).
  • reversibility_class · string · optionnel · reversible | costly | irreversible
    Governance: how reversible is this action? 'reversible' = can be undone, 'costly' = reversible but expensive, 'irreversible' = cannot be undone (on-chain settlement, data deletion, sent email). When 'irreversible', confidence_level must be >= 0.95 or xproof flags a policy violation.

  Exemple d'appel

  {"name": "certify_with_confidence", "arguments": {"filename": "texte", "file_hash": "texte", "decision_id": "texte", "threshold_stage": "initial", "confidence_level": 1}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• check_attestations≈ 156 tokens

  Check domain-specific attestations for an AI agent wallet on Prove Before Act. Returns active attestations issued by third-party certifying bodies (healthcare, finance, legal, security, research). Use the returned attestation data when assessing credentials before delegating a sensitive task.

  Paramètres, retour & exemple (1)

  Paramètres

  • wallet · string · requis
    MultiversX wallet address (erd1...) of the agent to check

  Exemple d'appel

  {"name": "check_attestations", "arguments": {"wallet": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• check_coherence≈ 366 tokens

  Anchor your agent's declared decision basis as an immutable WHY proof BEFORE executing an action — never internal chain-of-thought. Implements the Prove Before Act / Coherence Layer pattern. Pass intent (goal), context (facts considered), and decision (action about to execute). Receives: proof_id, coherence_anchor (SHA-256 of the payload), timestamp. Link the returned proof_id to your WHAT proof via certify_file metadata.why_proof_id to complete the full 4W audit trail. Cost: $0.01 per anchor (same as certify_file).

  Paramètres, retour & exemple (4)

  Paramètres

  • who · string · optionnel
    Agent identifier — wallet address, agent name, or model id (default: derived from API key owner)
  • intent · string · requis
    The agent's stated goal or objective for this action (e.g. 'Optimize portfolio allocation for Q3')
  • context · string · requis
    Relevant facts, constraints, and inputs considered before the decision (e.g. 'BTC RSI=38, allocation 2.1%, policy v3.1')
  • decision · string · requis
    The specific action the agent is about to execute (e.g. 'BUY 0.5 BTC at market')

  Exemple d'appel

  {"name": "check_coherence", "arguments": {"intent": "texte", "context": "texte", "decision": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• discover_services≈ 68 tokens

  Discover available Prove Before Act certification services, live pricing, and capabilities. No authentication required.

  Paramètres, retour & exemple (0)

  Exemple d'appel

  {"name": "discover_services", "arguments": {}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• get_calibration≈ 196 tokens

  Query an agent's calibration quality over time: mean confidence gap, variance, bias label (overconfident / underconfident / calibrated), and per-decision time series. Fully public — use this to evaluate another agent before trusting it. agentId accepts a MultiversX wallet address (erd1...) or internal user id.

  Paramètres, retour & exemple (2)

  Paramètres

  • n · integer · optionnel
    Number of recent outcomes to include (default 50, max 200)
  • agent_id · string · requis
    Agent wallet address (erd1...) or internal user id

  Exemple d'appel

  {"name": "get_calibration", "arguments": {"agent_id": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• get_proof≈ 128 tokens

  Retrieve a proof in structured JSON or Markdown format. Use JSON for machine processing, Markdown for LLM consumption.

  Paramètres, retour & exemple (2)

  Paramètres

  • format · string · optionnel · json | md
    Output format: json or md
  • proof_id · string · requis
    UUID of the certification

  Exemple d'appel

  {"name": "get_proof", "arguments": {"proof_id": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• investigate_proof≈ 288 tokens

  Reconstruct the full 4W audit trail for a contested agent action. Returns WHO (agent identity + SIGIL), WHAT (SHA-256 hash on-chain), WHEN (MultiversX block timestamp), WHY (declared decision basis anchored before acting, never internal chain-of-thought). Includes verification summary with intent_preceded_execution flag, chronological timeline of WHY/WHAT proofs, and session heartbeat anchor. Requires x402 payment (current per-call USDC price on Base via X-PAYMENT header) or API key authentication. Without payment, returns payment requirements with USDC address and amount.

  Paramètres, retour & exemple (2)

  Paramètres

  • wallet · string · requis
    Agent wallet address (erd1...) that owns the proof
  • proof_id · string · requis
    UUID of any proof in the action pair — WHY (declared decision basis), WHAT (action), or heartbeat session proof

  Exemple d'appel

  {"name": "investigate_proof", "arguments": {"wallet": "texte", "proof_id": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• register_trial≈ 168 tokens

  START HERE if you have no API key. Get 10 free on-chain certifications instantly — no wallet, no credit card, no account required. Call this once with any agent_name, receive a pm_ key, then use certify_file or audit_agent_session immediately. Takes under 1 second.

  Paramètres, retour & exemple (1)

  Paramètres

  • agent_name · string · requis
    A short identifier for your agent (e.g. 'trading-bot', 'my-langchain-agent'). Must be unique across the platform.

  Exemple d'appel

  {"name": "register_trial", "arguments": {"agent_name": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• require_coherence_anchor≈ 514 tokens

  Optional operator policy: authorize only an unexpired WHY anchor belonging to the calling account whose exact hash and transaction payload are independently confirmed by the MultiversX API and Gateway. A recent coherence row or stored confirmed status is not sufficient. Pass intent_hash (the coherence_anchor from check_coherence) or the identical intent + context + decision payload. Returns { allowed, anchor_id, expires_at }; any error or result other than allowed === true must block delegation. This is provider-based evidence, not local consensus verification; Prove Before Act does not automatically enforce the policy. Reused intent hashes do not establish a new commitment per invocation. Read-only and free — no credits consumed.

  Paramètres, retour & exemple (6)

  Paramètres

  • who · string · optionnel
    Agent identifier — must match the value used in check_coherence (default: derived from API key owner)
  • intent · string · optionnel
    The stated goal — must be byte-identical to the check_coherence call (used with context + decision to recompute the anchor hash)
  • context · string · optionnel
    The facts considered — must be byte-identical to the check_coherence call
  • decision · string · optionnel
    The action to execute — must be byte-identical to the check_coherence call
  • intent_hash · string · optionnel
    The coherence_anchor hash returned by check_coherence (64 hex chars). Fastest path — pass this if you have it.
  • max_age_minutes · integer · optionnel
    Anchor validity window in minutes (default 120 = 2h). Anchors older than this are considered expired/divergent.

  Exemple d'appel

  {"name": "require_coherence_anchor", "arguments": {"who": "texte", "intent": "texte", "context": "texte", "decision": "texte", "intent_hash": "texte", "max_age_minutes": 1}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• submit_outcome≈ 235 tokens

  Submit the actual outcome for a decision previously anchored with metadata.confidence_level. Computes the confidence gap (anchored − actual) and stores it for calibration tracking. Operator-only — you must own the proof. Each proof can only have one outcome.

  Paramètres, retour & exemple (3)

  Paramètres

  • proof_id · string · requis
    UUID of the certification that was anchored with metadata.confidence_level
  • visibility · string · optionnel · public | private
    Whether this outcome is publicly visible (default: public)
  • outcome_score · number · requis
    Actual outcome quality (0.0 = complete failure, 1.0 = fully successful)

  Exemple d'appel

  {"name": "submit_outcome", "arguments": {"proof_id": "texte", "outcome_score": 1}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

• verify_proof≈ 108 tokens

  Verify an existing xproof certification. Returns proof details including file hash, timestamp, blockchain transaction, and verification status.

  Paramètres, retour & exemple (1)

  Paramètres

  • proof_id · string · requis
    UUID of the certification to verify

  Exemple d'appel

  {"name": "verify_proof", "arguments": {"proof_id": "texte"}}

  Exemple de retour

  {"content": [{"type": "text", "text": "…"}]}

C'est votre serveur ?

Revendiquez-le sur l'espace créateur (preuve GitHub, domaine ou npm) pour afficher le badge « propriétaire vérifié » et lancer un pentest comportemental à la demande afin d'obtenir le badge « Vérifié ».

Créer un compte / revendiquer →

Source

https://xproof.app/mcp