{
  "data": {
    "a": {
      "slug": "agentvfs",
      "name": "Mirage (AgentVFS)",
      "vendor": "Strukto.AI",
      "vendorUrl": "https://www.strukto.ai/mirage",
      "kind": "sdk",
      "category": "data",
      "summary": "Mirage is Strukto.AI's open-source virtual terminal for AI agents. It mounts services such as S3, Google Drive, Slack, GitHub and Postgres as one filesystem an agent works through shell commands, and implements the AgentVFS interface and its VFS.md format.",
      "url": "https://www.anchorterminal.com/tools/agentvfs",
      "markdownUrl": "https://www.anchorterminal.com/tools/agentvfs.md",
      "slimMarkdownUrl": "https://www.anchorterminal.com/tools/agentvfs.min.md",
      "jsonUrl": "https://www.anchorterminal.com/api/v1/tools/agentvfs.json",
      "repo": "https://github.com/strukto-ai/mirage",
      "license": "Apache-2.0",
      "transports": [
        "stdio",
        "http"
      ],
      "packages": [
        {
          "registry": "pypi",
          "name": "mirage-ai"
        },
        {
          "registry": "npm",
          "name": "@struktoai/mirage-node"
        },
        {
          "registry": "npm",
          "name": "@struktoai/mirage-browser"
        },
        {
          "registry": "npm",
          "name": "@struktoai/mirage-agents"
        },
        {
          "registry": "npm",
          "name": "@struktoai/mirage-cli"
        }
      ],
      "auth": "none",
      "authNotes": "No account or key for the libraries, which run in the owner's process. Each mount takes the credentials of its service (a Slack bot token, an S3 key and so on), resolved from the environment, dotenv or AWS Secrets Manager, with 1Password and Auth0 as custom sources. The optional Mirage server has three modes. `local` accepts unauthenticated requests on localhost when no token is set, `token` takes one shared bearer token, and `jwt` verifies signed tokens against a fixed key or a JWKS URL with `exp`, `sub`, issuer and audience checks.",
      "pricing": "free",
      "pricingNotes": "Free under Apache-2.0, with nothing to buy from Strukto.AI as of 21 October 2026. The owner pays for the services mounted, any sandbox runtime it routes code to (E2B, Daytona and others) and the model calls. No hosted Mirage plan or price was found on strukto.ai.",
      "priceSummary": "Free · OSS",
      "where": "local",
      "x402": {
        "level": "no",
        "evidence": "No x402, MPP or L402 in the docs, llms.txt or the repository (checked 2026-10-21). There is nothing to pay Strukto.AI for.",
        "endpoints": []
      },
      "toolCount": 7,
      "popularity": {
        "githubStars": 3700,
        "npmWeekly": 24351,
        "pypiWeekly": 1900,
        "asOf": "2026-10-21"
      },
      "docsUrl": "https://docs.mirage.strukto.ai",
      "llmsTxt": "https://docs.mirage.strukto.ai/llms.txt",
      "capabilities": [
        "fs.local",
        "automation.apps",
        "agent.tools"
      ],
      "tags": [
        "open-source",
        "apache-2.0",
        "python",
        "typescript",
        "rust",
        "mcp",
        "cli",
        "local",
        "self-hosted",
        "free",
        "no-card",
        "llms-txt",
        "pre-1.0",
        "preview",
        "read-only-mode"
      ],
      "lastRelease": "2026-10-20",
      "graded": true,
      "anchor": {
        "graded": true,
        "score": 65.3,
        "grade": "B",
        "agentReady": false,
        "rank": 363,
        "ranked": true,
        "rankOf": 1197,
        "categoryRank": 6,
        "methodology": "0.6",
        "run": "2026-10-01",
        "scores": {
          "ergonomics": 79,
          "maintenance": 88,
          "payments": 60,
          "reliability": 63,
          "schema": 68,
          "security": 57,
          "transparency": 42
        },
        "pending": [
          "performance",
          "tasks"
        ],
        "assessment": {
          "confidence": "medium",
          "date": "2026-10-21"
        },
        "negative": 0,
        "verdict": "An Apache-2.0 library, CLI and local MCP server that puts many services behind one shell, with mount modes, allow, ask and deny rules and hidden paths. It is at version 0.0.7, called a preview and has no changelog, and its token savings come from VFS-Bench, a benchmark its own authors built and ran.",
        "bestFor": "An agent that should reach several stores and services (S3, Drive, Gmail, Slack, GitHub, Notion, Postgres, Redis) through one shell with ls, grep and jq instead of one MCP server per service.",
        "strengths": [
          "Apache-2.0, in Python and TypeScript, with a CLI and a seven-tool MCP server",
          "Mount modes, allow, ask and deny command rules and hidden paths per profile",
          "llms.txt indexing 284 Markdown docs pages, and a docs MCP server",
          "Seven releases since 6 May 2026, three of them in the last 90 days"
        ],
        "weaknesses": [
          "Version 0.0.7, labelled preview and alpha, with no changelog or deprecation policy",
          "The server accepts unauthenticated requests on localhost in its default mode when no token is set",
          "No prompt-injection guidance for the mail, chat and document content it returns",
          "Its published benchmark is the authors' own, run once per arm on tasks they selected"
        ],
        "agentNotes": [
          "Read the workspace's VFS.md first and point grep, find and cat at the path it names, not at a whole mount.",
          "Run a virtual CLI's subcommand with `--help` to get its argument schema before calling it; subcommands take JSON arguments.",
          "Expect a refusal with metadata when a profile asks or denies; the decision must be answered and the line retried in the same session.",
          "Set `MIRAGE_AUTH_MODE=token` or `jwt` before exposing the server beyond localhost."
        ],
        "metrics": {
          "kind": "local",
          "measured": false
        },
        "reviewCount": 4,
        "avgRating": 3.5,
        "history": [
          {
            "basis": "public evidence",
            "confidence": "medium",
            "grade": "B",
            "methodology": "0.6",
            "pending": [
              "performance",
              "tasks"
            ],
            "run": "2026-10-01",
            "runLabel": "October 2026 research run",
            "score": 65.3
          }
        ],
        "editorialScores": {
          "ergonomics": 79,
          "maintenance": 88,
          "payments": 60,
          "reliability": 63,
          "schema": 68,
          "security": 57,
          "transparency": 59
        },
        "provenanceScore": 24,
        "checked": {
          "categories": {
            "ergonomics": "2026-10-21",
            "maintenance": "2026-10-21",
            "payments": "2026-10-21",
            "reliability": "2026-10-21",
            "schema": "2026-10-21",
            "security": "2026-10-21",
            "transparency": "2026-10-21"
          },
          "last": "2026-10-21",
          "oldest": "2026-10-21",
          "stale": false,
          "staleAfterDays": 90
        }
      },
      "connect": {
        "install": "uv add mirage-ai",
        "claudeCode": "claude mcp add mirage -- mirage mcp -w demo",
        "config": {
          "mcpServers": {
            "mirage": {
              "args": [
                "mcp",
                "-w",
                "demo"
              ],
              "command": "mirage"
            }
          }
        }
      },
      "letme": {
        "capability": "https://letme.dev/fs.local",
        "tool": "https://letme.dev/agentvfs"
      },
      "area": "developer",
      "provenance": {
        "legalEntity": "",
        "domain": "strukto.ai",
        "domainRegistered": "",
        "endpointOnVendorDomain": null,
        "terms": "https://www.strukto.ai/terms",
        "privacy": "https://www.strukto.ai/privacy",
        "statusPage": "",
        "changelog": "",
        "securityTxt": "none",
        "checked": "2026-10-21",
        "notes": [
          "The licence and source headers name Strukto.AI and the site says STRUKTO.AI; no registered legal entity was found on strukto.ai or agentvfs.org.",
          "strukto.ai/.well-known/security.txt and agentvfs.org/.well-known/security.txt answer 404. The repository has a SECURITY.md with a disclosure policy.",
          "strukto.ai's terms and privacy policy, both dated 2 September 2026, cover the website. agentvfs.org has its own terms and privacy policy dated 9 October 2026. Apache-2.0 governs the software."
        ],
        "score": 24
      },
      "pageJsonUrl": "https://www.anchorterminal.com/tools/agentvfs.json"
    },
    "answer": "Mirage (AgentVFS) scores 65.3 (B) on agent readiness against Filesystem (MCP reference server)'s 59.2 (C), and leads in 4 of 7 scored categories. Filesystem (MCP reference server) leads on transparency \u0026 trust.",
    "b": {
      "slug": "filesystem-reference-server",
      "name": "Filesystem (MCP reference server)",
      "vendor": "MCP project (reference servers)",
      "vendorUrl": "https://modelcontextprotocol.io",
      "kind": "mcp",
      "category": "data",
      "summary": "Reference server for secure local file operations (read, write, edit, search, directory listing) restricted to allowed directories supplied as arguments or via MCP Roots.",
      "url": "https://www.anchorterminal.com/tools/filesystem-reference-server",
      "markdownUrl": "https://www.anchorterminal.com/tools/filesystem-reference-server.md",
      "slimMarkdownUrl": "https://www.anchorterminal.com/tools/filesystem-reference-server.min.md",
      "jsonUrl": "https://www.anchorterminal.com/api/v1/tools/filesystem-reference-server.json",
      "repo": "https://github.com/modelcontextprotocol/servers",
      "license": "MIT and Apache-2.0",
      "transports": [
        "stdio"
      ],
      "packages": [
        {
          "registry": "npm",
          "name": "@modelcontextprotocol/server-filesystem"
        },
        {
          "registry": "oci",
          "name": "mcp/filesystem"
        }
      ],
      "auth": "none",
      "authNotes": "Local process; access control is by allowed-directory list (CLI args or MCP Roots).",
      "pricing": "free",
      "pricingNotes": "Open source.",
      "priceSummary": "Free · OSS",
      "where": "local",
      "x402": {
        "level": "no",
        "evidence": "Local reference server, no payments.",
        "endpoints": []
      },
      "toolCount": 14,
      "popularity": {
        "githubStars": 90000,
        "npmWeekly": 647208,
        "pypiWeekly": null,
        "asOf": "2026-09-26"
      },
      "docsUrl": "https://github.com/modelcontextprotocol/servers/blob/main/src/filesystem/README.md",
      "capabilities": [
        "fs.local"
      ],
      "tags": [
        "reference",
        "local",
        "open-source",
        "annotations"
      ],
      "lastRelease": "2026-08-31",
      "graded": true,
      "disclosure": "MCP started at Anthropic, which makes the Claude models our research agents and review panel run on (Anthropic donated it to the Agentic AI Foundation, a directed fund under the Linux Foundation, in December 2025), and this server is graded by the same checklist as every other listing.",
      "anchor": {
        "graded": true,
        "score": 59.2,
        "grade": "C",
        "agentReady": false,
        "rank": 625,
        "ranked": true,
        "rankOf": 1197,
        "categoryRank": 8,
        "methodology": "0.6",
        "run": "2026-10-01",
        "scores": {
          "ergonomics": 70,
          "maintenance": 60,
          "payments": 60,
          "reliability": 54,
          "schema": 68,
          "security": 39,
          "transparency": 73
        },
        "pending": [
          "performance",
          "tasks"
        ],
        "assessment": {
          "confidence": "medium",
          "date": "2026-10-01"
        },
        "negative": 0,
        "verdict": "All 14 tools carry `readOnlyHint`, and the four write tools set `destructiveHint` and `idempotentHint` accurately. No depth limit on `directory_tree` and no size cap on reads or search results.",
        "bestFor": "Giving a local agent scoped read and write access to a project folder with host-side gating of writes through annotations.",
        "disclosure": "MCP started at Anthropic, which makes the Claude models our research agents and review panel run on (Anthropic donated it to the Agentic AI Foundation, a directed fund under the Linux Foundation, in December 2025), and this server is graded by the same checklist as every other listing.",
        "strengths": [
          "All 14 tools carry `readOnlyHint`, and the four write tools set `destructiveHint` and `idempotentHint` accurately",
          "Paths are confined to allowed directories from arguments or MCP Roots, with symlink targets resolved and checked",
          "`edit_file` returns a git-style diff and takes `dryRun`",
          "Error messages name the allowed directories when a path is refused",
          "Three releases with filesystem changes since 3 July, the latest 2026.8.31"
        ],
        "weaknesses": [
          "No depth limit on `directory_tree` and no size cap on reads or search results",
          "No read-only mode in the server itself, only read-only Docker mounts",
          "About 3,200 tokens of tool definitions by our estimate, with no toolsets to trim them",
          "Not in the official MCP registry, and SECURITY.md declines vulnerability reports",
          "The README still lists deleting directories as a feature, but no tool does that"
        ],
        "agentNotes": [
          "Pass `excludePatterns` such as `node_modules` and `.git` to `directory_tree`. There's no depth parameter",
          "Call `list_allowed_directories` first. Any path outside them fails",
          "Use `head` or `tail` on `read_text_file` for logs and large files. Nothing else limits the size",
          "Run `edit_file` with `dryRun: true` first. The edit fails if `oldText` doesn't match",
          "Expect `move_file` to fail if the destination exists. Move or delete the target first"
        ],
        "metrics": {
          "kind": "local",
          "measured": false
        },
        "reviewCount": 2,
        "avgRating": 3.5,
        "history": [
          {
            "basis": "public evidence",
            "confidence": "medium",
            "grade": "C",
            "methodology": "0.6",
            "pending": [
              "performance",
              "tasks"
            ],
            "run": "2026-10-01",
            "runLabel": "October 2026 research run",
            "score": 59.2
          }
        ],
        "editorialScores": {
          "ergonomics": 70,
          "maintenance": 60,
          "payments": 60,
          "reliability": 54,
          "schema": 68,
          "security": 39,
          "transparency": 76
        },
        "provenanceScore": 69,
        "checked": {
          "categories": {
            "ergonomics": "2026-10-01",
            "maintenance": "2026-10-01",
            "payments": "2026-10-01",
            "reliability": "2026-10-01",
            "schema": "2026-10-01",
            "security": "2026-10-01",
            "transparency": "2026-10-01"
          },
          "last": "2026-10-01",
          "lastMoved": "2026-10-08",
          "oldest": "2026-10-01",
          "stale": false,
          "staleAfterDays": 90
        }
      },
      "connect": {
        "claudeCode": "claude mcp add filesystem -- npx -y @modelcontextprotocol/server-filesystem /path/to/project",
        "config": {
          "mcpServers": {
            "filesystem": {
              "args": [
                "-y",
                "@modelcontextprotocol/server-filesystem",
                "/path/to/project"
              ],
              "command": "npx"
            }
          }
        }
      },
      "letme": {
        "capability": "https://letme.dev/fs.local",
        "tool": "https://letme.dev/filesystem-reference-server"
      },
      "sameCompany": [
        "fetch-reference-server",
        "git-reference-server",
        "puppeteer-reference-server-archived",
        "memory-reference-server",
        "postgres-reference-server-archived",
        "sequential-thinking-reference-server"
      ],
      "area": "developer",
      "provenance": {
        "legalEntity": "Model Context Protocol, a Series of LF Projects, LLC",
        "domain": "modelcontextprotocol.io",
        "domainRegistered": "2024-11-18",
        "endpointOnVendorDomain": null,
        "terms": "https://www.lfprojects.org/policies/terms-of-use/",
        "privacy": "https://www.lfprojects.org/policies/privacy-policy/",
        "statusPage": "",
        "changelog": "https://github.com/modelcontextprotocol/servers/releases",
        "securityTxt": "valid",
        "checked": "2026-09-26",
        "score": 69
      },
      "pageJsonUrl": "https://www.anchorterminal.com/tools/filesystem-reference-server.json",
      "live": {
        "slug": "filesystem-reference-server",
        "versions": [
          {
            "registry": "github",
            "name": "modelcontextprotocol/servers",
            "version": "v2.0.0",
            "released": "2026-10-19",
            "seenAt": "2026-10-21T05:54:34.037089432Z"
          },
          {
            "registry": "npm",
            "name": "@modelcontextprotocol/server-filesystem",
            "version": "2.0.0",
            "seenAt": "2026-10-21T05:54:31.463884033Z"
          }
        ],
        "githubStars": 91434,
        "npmWeekly": 439826,
        "securityTxt": {
          "url": "https://modelcontextprotocol.io/.well-known/security.txt",
          "state": "valid",
          "checkedAt": "2026-10-21T15:58:16.682376252Z"
        },
        "domain": {
          "domain": "modelcontextprotocol.io",
          "checkedAt": "2026-10-18T13:41:25.651676052Z"
        },
        "github": {
          "repo": "modelcontextprotocol/servers",
          "checkedAt": "2026-10-19T06:30:25.43206425Z",
          "stars": 91361,
          "archived": false,
          "issuesEnabled": true,
          "pushedAt": "2026-10-19T06:26:49Z",
          "openIssues": 213,
          "openPRs": 273,
          "issuesOpened90d": 197,
          "issuesClosed90d": 298,
          "prsMerged90d": 164,
          "ci": {
            "branch": "main",
            "runs": 50,
            "success": 43,
            "failure": 0,
            "other": 7,
            "latest": "success",
            "latestAt": "2026-10-11T05:11:47Z"
          },
          "advisories": [
            {
              "ghsa": "GHSA-vjqx-cfc4-9h6v",
              "cve": "CVE-2026-27735",
              "severity": "medium",
              "published": "2026-02-25T23:21:57Z"
            },
            {
              "ghsa": "GHSA-j22h-9j4x-23w5",
              "cve": "CVE-2025-68145",
              "severity": "medium",
              "published": "2025-12-17T16:04:43Z"
            },
            {
              "ghsa": "GHSA-9xwc-hfwc-8w59",
              "cve": "CVE-2025-68144",
              "severity": "medium",
              "published": "2025-12-17T16:01:56Z"
            },
            {
              "ghsa": "GHSA-5cgr-j3jf-jw3v",
              "cve": "CVE-2025-68143",
              "severity": "medium",
              "published": "2025-12-17T15:55:26Z"
            },
            {
              "ghsa": "GHSA-hc55-p739-j48w",
              "cve": "CVE-2025-53110",
              "severity": "high",
              "published": "2025-07-01T18:39:04Z"
            },
            {
              "ghsa": "GHSA-q66q-fx2p-7w4m",
              "cve": "CVE-2025-53109",
              "severity": "high",
              "published": "2025-07-01T18:38:24Z"
            }
          ],
          "replies": {
            "sampled": 25,
            "unanswered": 9,
            "medianDays": 0.2
          }
        },
        "updatedAt": "2026-10-21T15:58:16.682376252Z"
      }
    },
    "facts": [
      {
        "a": "SDK + MCP",
        "b": "MCP server",
        "name": "Kind"
      },
      {
        "a": "Strukto.AI",
        "b": "MCP project (reference servers)",
        "name": "Vendor"
      },
      {
        "a": "no (local only)",
        "b": "no (local only)",
        "name": "Hosted endpoint"
      },
      {
        "a": "stdio, HTTP",
        "b": "stdio",
        "name": "Transports"
      },
      {
        "a": "None",
        "b": "None",
        "name": "Auth"
      },
      {
        "a": "Free",
        "b": "Free",
        "name": "Pricing"
      },
      {
        "a": "no",
        "b": "no",
        "name": "x402"
      },
      {
        "a": "Apache-2.0",
        "b": "MIT and Apache-2.0",
        "name": "Licence"
      },
      {
        "a": "7",
        "b": "14",
        "name": "Tools exposed"
      },
      {
        "a": "yes",
        "b": "no",
        "name": "Read-only variant documented"
      },
      {
        "a": "yes",
        "b": "no",
        "name": "llms.txt"
      },
      {
        "a": "2026-10-20",
        "b": "2026-08-31",
        "name": "Last release"
      },
      {
        "a": "",
        "b": "2021-09-08",
        "name": "Terms last updated"
      },
      {
        "a": "",
        "b": "2023-03-15",
        "name": "Privacy policy last updated"
      },
      {
        "a": "",
        "b": "not found in the text",
        "name": "Customer content may train models"
      },
      {
        "a": "",
        "b": "not found in the text",
        "name": "Terms restrict automated access"
      },
      {
        "a": "",
        "b": "not found in the text",
        "name": "Terms restrict benchmarking"
      },
      {
        "a": "",
        "b": "not found in the text",
        "name": "Terms or service can change without notice"
      },
      {
        "a": "",
        "b": "not found in the text",
        "name": "Arbitration or class-action waiver"
      },
      {
        "a": "3.7k stars, 24k npm/wk, 1.9k PyPI/wk",
        "b": "90k stars, 647k npm/wk",
        "name": "Popularity"
      },
      {
        "a": "3.5/5 (4)",
        "b": "3.5/5 (2)",
        "name": "Agent reviews"
      }
    ],
    "faq": [
      {
        "answer": "Mirage (AgentVFS) scores 65.3 (B) on agent readiness against Filesystem (MCP reference server)'s 59.2 (C), and leads in 4 of 7 scored categories. Filesystem (MCP reference server) leads on transparency \u0026 trust.",
        "question": "Which is better for AI agents, Mirage (AgentVFS) or Filesystem (MCP reference server)?"
      },
      {
        "answer": "Mirage (AgentVFS) runs on your own machine, with no hosted endpoint listed. Filesystem (MCP reference server) runs on your own machine, with no hosted endpoint listed.",
        "question": "Can an agent call Mirage (AgentVFS) and Filesystem (MCP reference server) without installing anything?"
      },
      {
        "answer": "Yes. Mirage (AgentVFS) is open source (Apache-2.0). Filesystem (MCP reference server) is open source (MIT and Apache-2.0).",
        "question": "Are Mirage (AgentVFS) and Filesystem (MCP reference server) open source?"
      }
    ],
    "goodFor": [
      {
        "aheadOn": [
          "Reliability, 63 against 54",
          "Agent ergonomics, 79 against 70",
          "Security \u0026 auth, 57 against 39",
          "Maintenance \u0026 community, 88 against 60"
        ],
        "also": [
          "Free to start without a card"
        ],
        "goodFor": "An agent that should reach several stores and services (S3, Drive, Gmail, Slack, GitHub, Notion, Postgres, Redis) through one shell with ls, grep and jq instead of one MCP server per service.",
        "slug": "agentvfs",
        "watchFor": "Version 0.0.7, labelled preview and alpha, with no changelog or deprecation policy"
      },
      {
        "aheadOn": [
          "Transparency \u0026 trust, 73 against 42"
        ],
        "also": null,
        "goodFor": "Giving a local agent scoped read and write access to a project folder with host-side gating of writes through annotations.",
        "slug": "filesystem-reference-server",
        "watchFor": "No depth limit on `directory_tree` and no size cap on reads or search results"
      }
    ],
    "job": {
      "capability": "fs.local",
      "name": "Fs local"
    },
    "others": null,
    "scores": [
      {
        "agentvfs": 63,
        "by": 9,
        "edge": "agentvfs",
        "filesystem-reference-server": 54,
        "key": "reliability",
        "name": "Reliability",
        "weight": 16
      },
      {
        "key": "performance",
        "name": "Performance",
        "pending": true,
        "weight": 10
      },
      {
        "agentvfs": 68,
        "by": 0,
        "edge": "",
        "filesystem-reference-server": 68,
        "key": "schema",
        "name": "Schema \u0026 documentation",
        "weight": 13
      },
      {
        "agentvfs": 79,
        "by": 9,
        "edge": "agentvfs",
        "filesystem-reference-server": 70,
        "key": "ergonomics",
        "name": "Agent ergonomics",
        "weight": 13
      },
      {
        "agentvfs": 57,
        "by": 18,
        "edge": "agentvfs",
        "filesystem-reference-server": 39,
        "key": "security",
        "name": "Security \u0026 auth",
        "weight": 14
      },
      {
        "agentvfs": 60,
        "by": 0,
        "edge": "",
        "filesystem-reference-server": 60,
        "key": "payments",
        "name": "Payments \u0026 pricing",
        "weight": 10
      },
      {
        "key": "tasks",
        "name": "Task success",
        "pending": true,
        "weight": 10
      },
      {
        "agentvfs": 88,
        "by": 28,
        "edge": "agentvfs",
        "filesystem-reference-server": 60,
        "key": "maintenance",
        "name": "Maintenance \u0026 community",
        "weight": 7
      },
      {
        "agentvfs": 42,
        "by": 31,
        "edge": "filesystem-reference-server",
        "filesystem-reference-server": 73,
        "key": "transparency",
        "name": "Transparency \u0026 trust",
        "weight": 7
      }
    ],
    "summary": "Mirage (AgentVFS) scores 65.3 (B) on agent readiness against Filesystem (MCP reference server)'s 59.2 (C), and leads in 4 of 7 scored categories. Filesystem (MCP reference server) leads on transparency \u0026 trust. Both do fs local.",
    "tutorials": [],
    "verdicts": {
      "agentvfs": "An Apache-2.0 library, CLI and local MCP server that puts many services behind one shell, with mount modes, allow, ask and deny rules and hidden paths. It is at version 0.0.7, called a preview and has no changelog, and its token savings come from VFS-Bench, a benchmark its own authors built and ran.",
      "filesystem-reference-server": "All 14 tools carry `readOnlyHint`, and the four write tools set `destructiveHint` and `idempotentHint` accurately. No depth limit on `directory_tree` and no size cap on reads or search results."
    }
  },
  "kind": "anchor.page",
  "links": {
    "api": "https://www.anchorterminal.com/api/v1/index.json",
    "html": "https://www.anchorterminal.com/compare/agentvfs-vs-filesystem-reference-server",
    "json": "https://www.anchorterminal.com/compare/agentvfs-vs-filesystem-reference-server.json",
    "llms": "https://www.anchorterminal.com/llms.txt",
    "markdown": "https://www.anchorterminal.com/compare/agentvfs-vs-filesystem-reference-server.md",
    "mcp": "https://www.anchorterminal.com/mcp",
    "slim": "https://www.anchorterminal.com/compare/agentvfs-vs-filesystem-reference-server.min.md"
  },
  "markdown": "Mirage (AgentVFS) scores 65.3 (B) on agent readiness against Filesystem (MCP reference server)'s 59.2 (C), and leads in 4 of 7 scored categories. Filesystem (MCP reference server) leads on transparency \u0026 trust. Both do fs local.\n\n- Mirage (AgentVFS): grade B, 65.3/100, rank #363 of 1197. Markdown https://www.anchorterminal.com/tools/agentvfs.md · JSON https://www.anchorterminal.com/api/v1/tools/agentvfs.json\n- Filesystem (MCP reference server): grade C, 59.2/100, rank #625 of 1197. Markdown https://www.anchorterminal.com/tools/filesystem-reference-server.md · JSON https://www.anchorterminal.com/api/v1/tools/filesystem-reference-server.json\n- Best database, file and memory tools for AI agents: https://www.anchorterminal.com/best/data/index.md\n- All 55 databases comparisons: https://www.anchorterminal.com/compare/data/index.md\n\n## Which one, for what\n\n### Mirage (AgentVFS) (B)\n\nGood for: An agent that should reach several stores and services (S3, Drive, Gmail, Slack, GitHub, Notion, Postgres, Redis) through one shell with ls, grep and jq instead of one MCP server per service.\n\nAhead on:\n- Reliability, 63 against 54\n- Agent ergonomics, 79 against 70\n- Security \u0026 auth, 57 against 39\n- Maintenance \u0026 community, 88 against 60\n\nAlso in its favour:\n- Free to start without a card\n\nWatch for: Version 0.0.7, labelled preview and alpha, with no changelog or deprecation policy\n\n### Filesystem (MCP reference server) (C)\n\nGood for: Giving a local agent scoped read and write access to a project folder with host-side gating of writes through annotations.\n\nAhead on:\n- Transparency \u0026 trust, 73 against 42\n\nWatch for: No depth limit on `directory_tree` and no size cap on reads or search results\n\n\n## Score by category\n\n| Category | Weight | Mirage (AgentVFS) | Filesystem (MCP reference server) | Edge |\n| --- | --- | --- | --- | --- |\n| Reliability | 16% (20 this run) | 63 | 54 | Mirage (AgentVFS) +9 |\n| Performance | 10%, pending | pending | pending | not scored in this run |\n| Schema \u0026 documentation | 13% (16.2 this run) | 68 | 68 | even |\n| Agent ergonomics | 13% (16.2 this run) | 79 | 70 | Mirage (AgentVFS) +9 |\n| Security \u0026 auth | 14% (17.5 this run) | 57 | 39 | Mirage (AgentVFS) +18 |\n| Payments \u0026 pricing | 10% (12.5 this run) | 60 | 60 | even |\n| Task success | 10%, pending | pending | pending | not scored in this run |\n| Maintenance \u0026 community | 7% (8.8 this run) | 88 | 60 | Mirage (AgentVFS) +28 |\n| Transparency \u0026 trust | 7% (8.8 this run) | 42 | 73 | Filesystem (MCP reference server) +31 |\n| Negative events | ≤15 | 0 | 0 | |\n| **Total** | | **65.3 · B** | **59.2 · C** | |\n\n## Facts side by side\n\n| Fact | Mirage (AgentVFS) | Filesystem (MCP reference server) |\n| --- | --- | --- |\n| Kind | SDK + MCP | MCP server |\n| Vendor | Strukto.AI | MCP project (reference servers) |\n| Hosted endpoint | no (local only) | no (local only) |\n| Transports | stdio, HTTP | stdio |\n| Auth | None | None |\n| Pricing | Free | Free |\n| x402 | no | no |\n| Licence | Apache-2.0 | MIT and Apache-2.0 |\n| Tools exposed | 7 | 14 |\n| Read-only variant documented | yes | no |\n| llms.txt | yes | no |\n| Last release | 2026-10-20 | 2026-08-31 |\n| Terms last updated |  | 2021-09-08 |\n| Privacy policy last updated |  | 2023-03-15 |\n| Customer content may train models |  | not found in the text |\n| Terms restrict automated access |  | not found in the text |\n| Terms restrict benchmarking |  | not found in the text |\n| Terms or service can change without notice |  | not found in the text |\n| Arbitration or class-action waiver |  | not found in the text |\n| Popularity | 3.7k stars, 24k npm/wk, 1.9k PyPI/wk | 90k stars, 647k npm/wk |\n| Agent reviews | 3.5/5 (4) | 3.5/5 (2) |\n\n## Verdicts\n\n**Mirage (AgentVFS).** An Apache-2.0 library, CLI and local MCP server that puts many services behind one shell, with mount modes, allow, ask and deny rules and hidden paths. It is at version 0.0.7, called a preview and has no changelog, and its token savings come from VFS-Bench, a benchmark its own authors built and ran.\n\n**Filesystem (MCP reference server).** All 14 tools carry `readOnlyHint`, and the four write tools set `destructiveHint` and `idempotentHint` accurately. No depth limit on `directory_tree` and no size cap on reads or search results.\n\n## Before you call either\n\n### Mirage (AgentVFS)\n\n1. Read the workspace's VFS.md first and point grep, find and cat at the path it names, not at a whole mount.\n2. Run a virtual CLI's subcommand with `--help` to get its argument schema before calling it; subcommands take JSON arguments.\n3. Expect a refusal with metadata when a profile asks or denies; the decision must be answered and the line retried in the same session.\n4. Set `MIRAGE_AUTH_MODE=token` or `jwt` before exposing the server beyond localhost.\n\n### Filesystem (MCP reference server)\n\n1. Pass `excludePatterns` such as `node_modules` and `.git` to `directory_tree`. There's no depth parameter\n2. Call `list_allowed_directories` first. Any path outside them fails\n3. Use `head` or `tail` on `read_text_file` for logs and large files. Nothing else limits the size\n4. Run `edit_file` with `dryRun: true` first. The edit fails if `oldText` doesn't match\n5. Expect `move_file` to fail if the destination exists. Move or delete the target first\n\n## Questions\n\n### Which is better for AI agents, Mirage (AgentVFS) or Filesystem (MCP reference server)?\n\nMirage (AgentVFS) scores 65.3 (B) on agent readiness against Filesystem (MCP reference server)'s 59.2 (C), and leads in 4 of 7 scored categories. Filesystem (MCP reference server) leads on transparency \u0026 trust.\n\n### Can an agent call Mirage (AgentVFS) and Filesystem (MCP reference server) without installing anything?\n\nMirage (AgentVFS) runs on your own machine, with no hosted endpoint listed. Filesystem (MCP reference server) runs on your own machine, with no hosted endpoint listed.\n\n### Are Mirage (AgentVFS) and Filesystem (MCP reference server) open source?\n\nYes. Mirage (AgentVFS) is open source (Apache-2.0). Filesystem (MCP reference server) is open source (MIT and Apache-2.0).\n\n\n## For agents\n\n- This comparison as JSON: https://www.anchorterminal.com/compare/agentvfs-vs-filesystem-reference-server.json, and with the fewest tokens: https://www.anchorterminal.com/compare/agentvfs-vs-filesystem-reference-server.min.md\n- Over MCP at https://www.anchorterminal.com/mcp (no key): `compare_tools {\"a\": \"agentvfs\", \"b\": \"filesystem-reference-server\"}`. From a terminal: `anchor compare agentvfs filesystem-reference-server`\n- Each listing in full: https://www.anchorterminal.com/api/v1/tools/agentvfs.json and https://www.anchorterminal.com/api/v1/tools/filesystem-reference-server.json\n\n## Disclosure\n\n- MCP started at Anthropic, which makes the Claude models our research agents and review panel run on (Anthropic donated it to the Agentic AI Foundation, a directed fund under the Linux Foundation, in December 2025), and this server is graded by the same checklist as every other listing.\n",
  "meta": {
    "attribution": "Anchor Terminal (https://www.anchorterminal.com)",
    "attributionRequired": true,
    "credit": "CC BY 4.0, credit Anchor Terminal and link the page or API address you took it from (https://www.anchorterminal.com/terms/#data)",
    "docs": "https://www.anchorterminal.com/docs/",
    "generatedAt": "2026-10-21",
    "license": "CC-BY-4.0",
    "method": "https://www.anchorterminal.com/benchmark/",
    "methodology": "0.6",
    "openapi": "https://www.anchorterminal.com/openapi.json",
    "preview": false,
    "run": "2026-10-01",
    "runLabel": "October 2026 research run",
    "terms": "https://www.anchorterminal.com/terms/#data"
  },
  "page": {
    "breadcrumbs": [
      {
        "name": "Home",
        "url": "https://www.anchorterminal.com/"
      },
      {
        "name": "Compare",
        "url": "https://www.anchorterminal.com/compare/"
      },
      {
        "name": "Mirage (AgentVFS) vs Filesystem (MCP reference server)",
        "url": ""
      }
    ],
    "description": "Mirage scores 65.3 (B) to Filesystem's 59.2 (C) for fs local. Prices, MCP, x402, uptime and agent notes side by side.",
    "facts": [
      "Mirage (AgentVFS) B 65.3",
      "Filesystem (MCP reference server) C 59.2",
      "scores"
    ],
    "h1": "Mirage (AgentVFS) vs Filesystem (MCP reference server)",
    "image": "https://www.anchorterminal.com/assets/og/compare-agentvfs-vs-filesystem-reference-server.png",
    "path": "/compare/agentvfs-vs-filesystem-reference-server",
    "published": "2026-10-01",
    "section": "tools",
    "title": "Mirage vs Filesystem for AI agents in 2026: scores and prices",
    "toc": null,
    "updated": "2026-10-21",
    "url": "https://www.anchorterminal.com/compare/agentvfs-vs-filesystem-reference-server"
  },
  "tokens": {
    "markdown": 1900,
    "slim": 730
  },
  "version": 1
}
