$ timeahead.in
/ servers/npm/@agent-infra/mcp-server-filesystem
npm

@agent-infra/mcp-server-filesystem

MCP server for filesystem access

36k stars323/wkupdated 24d agogithub ↗
66fair
▣ Overview

What it does

This MCP server exposes filesystem operations as tools for use by Claude and other AI agents. It provides 12 tools for reading, writing, listing, and managing files and directories, enabling agents to interact with the file system as part of automated workflows. The server integrates into the Agent TARS ecosystem, a multimodal AI agent stack designed for automation, task execution, and agent orchestration. In addition to filesystem operations, the server makes network calls, though the specific networking purposes are not detailed in available documentation and should be verified during deployment.

Who it's for

Developers building autonomous agents that require controlled filesystem access. This is primarily useful for teams deploying Agent TARS or similar agentic frameworks where language models need to interact with files and directories as part of automated workflows, configuration management, or state persistence. Also suitable for agents that generate reports, logs, or other file-based output.

Common use cases

  • Enabling agents to read configuration files, documents, and knowledge bases as context for reasoning and decision-making
  • Allowing agents to write generated content, analysis, logs, and reports to disk
  • Integrating file operations into Agent TARS automation workflows and task pipelines
  • Filesystem-based state and memory management for long-running agentic processes
  • File exploration and navigation as part of agent-driven system administration or DevOps tasks

Setup pitfalls

  • Security scanner detected 3 hardcoded secrets in the codebase—review the source, rotate any exposed credentials, and clean the repository history before production use
  • High-risk classification requires strict sandboxing; filesystem write access should be limited to safe, isolated directories via environment configuration or container restrictions
  • No CI pipeline configured—the package lacks automated testing, linting, and security checks; audit and manually validate each update before upgrading
  • The server makes network calls; audit outbound traffic and verify all contacted endpoints during security review
3 credentials detected in repository history via Gitleaks
▣ Score BreakdownMCPScore = Σ(raw × weight)
DimensionRawWeighted
Security
35%
40
14.0
Freshness
25%
100
25.0
Adoption
20%
70
14.0
Quality
10%
80
8.0
Trust
10%
50
5.0
Total
66.0
⚿ Capabilities & Risk Explainer
fs readfs writenetworkexecevalsecrets
◆ Risk level: high· 12 tools · auth: API key
fs read + fs write + network + exec + eval + secrets active — can execute code, access credentials, and make external network calls.
Tool nameDescriptionDestructive?
web_search✓ no
read_file✓ no
read_multiple_files✓ no
write_file⚠ yes
edit_file✓ no
+7 more tools
create_directory✓ no
list_directory✓ no
directory_tree✓ no
move_file✓ no
search_files✓ no
get_file_info✓ no
list_allowed_directories✓ no
⚙ Install config
Claude Desktop · Cursor · Windsurf · VS Code (Copilot) · Claude Code
add to your MCP client config:
{
  "mcpServers": {
    "agent-inframcp-filesystem": {
      "command": "npx",
      "args": [
        "-y",
        "@agent-infra/mcp-server-filesystem"
      ]
    }
  }
}
📈 Score historylast 33 snapshots
4/30/20266/11/2026 · 33 snapshots
⚙ Maintenance health
56/ 100 · is this project alive?
contributors (1y)22
top contributor share78%
releases (1y)15
last release219d ago
ci✓ passing
⛁ Raw data
weekly downloads323
github stars36k
forks4k
open issues404
license✓ present
readme length16113 chars
last publish0d ago
last commit24d ago
last updated1d ago
install verified✓ pass · 23d ago
owner of this server? claim your listing to get a verified badgeclaim →
🔔 Score drop alerts
get notified by email when this server's score drops 5+ points