Skip to content
Philidor Labs logo

Philidor Labs MCP server

Which MCP servers are Philidor Labs's own, from the official registry and from asking its own domain, whether they answer, and what they offer. A program checked each of these, nobody copied them from a product page.

read 2026-10-26

← Back to Philidor Labs

Whose server it is
  • Registry nameio.philidor/defi-vaultsverified in the MCP registry

The registry name sits under philidor.io. The registry only lets a publisher use that namespace after proving control of the domain through DNS or HTTP.

When we asked itchecked 2026-10-26
  • Stateanswerschecked 2026-10-26
  • Tools24
  • Server calls itselfphilidor-defi-vaults
  • Server version1.0.0
  • Protocol version2025-06-18

On 2026-10-26 the server answered initialize and tools/list without a sign-in. First answer in 370 ms.

Should an agent connect itConnect it
  • Its tools do every use case we checked: Score and monitor vaults on a 0-10 risk scale; Check portfolio exposure and limits against an allocation policy.

Once connected

  • For "Score and monitor vaults on a 0-10 risk scale", the likely first call is search_vaults.
  • For "Check portfolio exposure and limits against an allocation policy", the likely first call is check_policy.

Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.

Measured dailydaily checks: 10, the first on 2026-10-26
  • INITE MCP score88 of 100answers 100 · does the job 100 · definitions 85 · context 87 · use 19 · connects 100
  • What an agent can do with it2 of 2 use cases
  • Use5 GitHub stars
  • Answered100% of days10 of 10 daily checks
  • Median first answer532 ms
  • Space in an agent's context≈ 4,778 tokensthe definitions of all its tools (24), which an agent carries on every turn
  • Tool definitionsB · 85 of 100
  • Typed results0 of 24 toolsdeclared outputSchema: results come typed, read once
  • Parameters without a description: in 4 of 24
  • Not saying whether it only reads or can destroy: 18 of 24

What an agent can do with it

  • can do · Score and monitor vaults on a 0-10 risk scale
  • can do · Check portfolio exposure and limits against an allocation policy
  • not sure · Generate signed risk reports for allocation decisions

Each of the product's use cases, put to Jev against this server's tool list: can an agent do it with these tools?

Worth a look

None of the five warning signs in these definitions.

A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (25%), doing the product's job (25%), definitions (20%), context (10%), use (10%) and how readily a client connects (10%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →

Tools it offers24 tools · read 2026-10-26
  • search_vaults

    Search and filter DeFi vaults by chain, protocol, asset, risk tier, TVL, and more. Returns a paginated list with risk scores and APR.

  • get_vault

    Get detailed information about a specific DeFi vault including risk breakdown, recent events, and historical snapshots. Lookup by ID or by network + address.

  • get_vault_risk_breakdown

    Get a detailed breakdown of a vault's risk vectors: Asset Composition, Platform and Strategy, Control and Governance, and History scores with sub-metrics.

  • list_markets

    List lending markets: the market a depositor actually picks (Aave/Spark pool, Aave V4 spoke, Compound Comet, Morpho Blue pair, or Kamino K-Lend market). Aave V4 spokes that share a liquidity hub are grouped under a composed parent…

  • get_market

    Get one lending market with every reserve it contains: supplied, supply APR, borrowed, borrow APR, utilization and risk tier per reserve. Spoke ids look like aave-v4-10-etherfi-cash. Aave V4 hub parents (aave-v4-1-hub-global-dollar) are…

  • get_market_events

    Published risk events for ONE lending market: incidents, bad debt, liquidation cascades (e.g. mass liquidations on a Morpho Blue market). Market ids look like morpho-v1-1-pt-reusd-10dec2026-usdc-91-5. Aave V4 hub ids…

  • check_loop_venue

    Underwrite a collateral/debt loop on ONE lending market: score, tier, utilization, borrow APR, deposit status, and recent incidents for both reserves. Does not compute health factor and does not prepare txs. Call this before a protocol…

  • list_ondo_assets

    Search the paginated Ondo Global Markets catalog: tokenized securities, portfolios and exact issuer-reported deployments. Includes API-only and CSV-only entries, freshness and identity conflicts. This is not a list of verified risk scores.

  • get_ondo_asset

    Get an exact Ondo token: issuer market observations, portfolio constituents, source freshness and chain/address assessment links. Prices are display-only; constituents have no inferred weights or inherited ratings.

  • get_ondo_asset_stream

    Get latest retained Ondo display prices and minute OHLC. Preserve source/receipt ages, evaluatedAt, conflicts and collector gaps; fresh is a 120-second policy at evaluatedAt, not an execution guarantee. Missing values are not zero. This…

  • get_ondo_asset_data

    Get Ondo OHLC, dividends, shares-multiplier history and per-chain token supply. Family availability and source clocks are explicit. Valid windows: 1min/1day, 5min/1day, 15min/1day, 1hour/1month, 4hour/1month, 12hour/3month, or 1day with…

  • get_native_asset_assessment

    Get the current effective score and signed reviewed base score for an exact Aptos, Sui, Noble, Stellar or XRP Ledger token. Preserve identifier case. A missing assessment is not a zero score. Evidence and signatures require the caller’s…

  • get_native_asset_financials

    Get public issuer financials for an exact Aptos, Sui, Noble, Stellar or XRP Ledger token. Preserve identifier case. Each metric states its unit, fund/global/deployment scope, source and freshness; null means unavailable, not zero. Fund…

  • compare_vaults

    Compare 2-3 DeFi vaults side-by-side on TVL, APR, risk score, risk tier, audited status, and more.

  • find_safest_vaults

    Find the highest risk-scored DeFi vaults, optionally filtered by asset, chain, or minimum TVL. Returns the top 10 audited vaults sorted by Philidor risk score (higher = lower assessed risk; scores are not a safety guarantee).

  • get_protocol_info

    Get detailed information about a DeFi protocol including TVL, vault count, versions, auditors, and security incidents.

  • get_curator_info

    Get detailed information about a vault curator including their managed vaults, TVL, chain distribution, and performance.

  • get_market_overview

    Get a high-level overview of the DeFi vault market: total TVL, vault count, risk distribution, and TVL by protocol.

  • explain_risk_score

    Explain what a Philidor risk score means, including the tier (Prime/Core/Edge), how it is calculated, and what the thresholds are.

  • list_vaults_with_incidents

    List all vaults that had a recent critical incident (last 365 days). Critical = severity Critical or incident_severity major. Results are sorted by TVL descending, then by time since event ascending (most recent first).

  • check_policy

    State whether the positions you supply conform to the limits and constraints you supply. Calls POST /v1/graph/check and returns the API compliance attestation (attestation de conformité): compliance (compliant or non_compliant)…

  • list_baskets

    Returns the structured access refusal for the baskets family. Does not return constituents or weights.

  • list_rwa

    Returns the structured access refusal for the rwa family. Does not return coverage.

  • get_oracles

    Returns the structured access refusal for the oracles family. Does not return freshness.

How to connect

Add the server to your MCP client config. Claude Desktop, Cursor and most other clients read this shape:

{
  "mcpServers": {
    "defi-vaults": {
      "type": "http",
      "url": "https://mcp.philidor.io/api/mcp"
    }
  }
}

Add it

Claude Code, in a terminal

claude mcp add --transport http philidor https://mcp.philidor.io/api/mcp

Add to CursorAdd to VS Code

claude.ai, ChatGPT and other clients: add https://mcp.philidor.io/api/mcp as a custom connector in the settings.

Each line is something a program found when it went looking: the registry entry, the server's own answer, a spec that parses. A missing line means we did not find it, not that it does not exist.