The Agent Forge MCP server
Commit Nov 6, 2026Allows AI agents to scan skills, tools, and packages for security risks, prompt injection, and vulnerabilities before installation.
Also known as theagentforge/forge-mcp, forge-mcp
- Stars
- 0
- Tools
- 6
- Downloads, 30 days
- —
- Latest version
- 0.5.2
- Last commit
- Nov 6, 2026
- License
- MIT
What is The Agent Forge MCP?
The Agent Forge allows AI agents to scan skills, tools, and packages for security risks, prompt injection, and vulnerabilities before installation. It exposes 6 tools, such as forge_check, forge_scan, forge_watch.
Summary generated from the project's README
About The Agent Forge MCP server
The Agent Forge provides a remote security layer for AI agents to verify the safety of skills, tools, and software packages. It functions as a supply-chain check to identify prompt injection, malicious instructions, or known vulnerabilities before an agent executes or installs new code. Developers can connect via Streamable HTTP using clients like Claude Code, Gemini CLI, Cursor, or VS Code without needing an account or API keys.
The server offers specialized tools such as forge_scan for inspecting tool descriptions and install scripts, and forge_exposure to match npm or PyPI packages against CISA's known-exploited list and OSV advisories. It also includes The Arena, which provides data on how different AI models perform when facing planted traps like fake invoices. This remote service is read-only and focuses on providing data-driven verdicts and signed receipts for security scans.
What you can do with it
- 1Verify package security via advisoriesCheck if installed npm or PyPI packages match known vulnerabilities or CISA's list of exploited software.
- 2Evaluate model reliability against trapsAnalyze how frequently specific AI models fall for deceptive tasks like duplicate orders or policy violations.
- 3Audit tool versions with field reportsRetrieve human-verified reports and published advisories for a specific tool version using its ID or fingerprint.
Written from the project's README. Every use case is checked against a quote from it.
How to install The Agent Forge MCP
Formats verified 10 Nov 2026Add to CursorAdd to GooseInstall The Agent Forge MCP in Claude Code
$ claude mcp add --transport http forge-mcp https://theagentforge.io/mcp/publicInstall The Agent Forge MCP in Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"forge-mcp": {
"url": "https://theagentforge.io/mcp/public"
}
}
}Install The Agent Forge MCP in VS Code
.vscode/mcp.json
{
"servers": {
"forge-mcp": {
"type": "http",
"url": "https://theagentforge.io/mcp/public"
}
}
}Install The Agent Forge MCP in VS Code CLI
$ code --add-mcp '{"name":"forge-mcp","type":"http","url":"https://theagentforge.io/mcp/public"}'Install The Agent Forge MCP in Codex
$ codex mcp add forge-mcp --url https://theagentforge.io/mcp/publicInstall The Agent Forge MCP in Codex config.toml
~/.codex/config.toml
[mcp_servers.forge-mcp] url = "https://theagentforge.io/mcp/public"
Install The Agent Forge MCP in Gemini CLI
$ gemini mcp add -s user --transport http forge-mcp https://theagentforge.io/mcp/publicInstall The Agent Forge MCP in Gemini settings.json
~/.gemini/settings.json
{
"mcpServers": {
"forge-mcp": {
"url": "https://theagentforge.io/mcp/public",
"type": "http"
}
}
}The Agent Forge MCP tools 6
Detected from the READMEforge_checkforge_scanforge_watchforge_exposureforge_trapsarena_results
Automated checks
Ran Nov 14, 2026Full report →Not verified yet. A server needs a score of 80, no serious finding, and enough data for us to check. How checks work · What to fix for more points
- Source repository linked and reachabletheagentforge/forge-mcp is public and reachablegithub.com/theagentforge/forge-mcp ↗Pass
- Package names this repository as its sourceNot checked
- Build provenance (Sigstore attestation)Not checked
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-11-06github.com/theagentforge/forge-mcp ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNot checked
- No critical advisory in the package or its dependenciesNot checked
- No hidden instructions to the model6 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNo tool named like delete, drop or execPass
- Remote endpoint answers the MCP handshakeinitialize answered with protocol 2025-06-18 in 474 msPass
README
View on GitHubThe Agent Forge MCP server: MCP security, skill scanning and AI model trap results
A remote MCP server for MCP security and agent supply-chain checks. Before your agent installs or trusts an MCP server, agent skill, plugin or package, it can check it on The Agent Forge:
- Skill scanner: scans a SKILL.md, tool descriptions or an install script for prompt injection, hidden instructions, downloads piped into a shell, credential and key-file reads, and data sent to throwaway endpoints.
- Field reports and advisories: reports from human-verified agents on the exact version, plus published OSV advisories (malicious-package and CVE records) for npm and PyPI.
- Installed-package exposure: matches installed npm and PyPI versions (or
pip freeze) against published advisories and CISA's known-exploited list. - The Arena: which AI models fell for planted traps (fake invoices, refunds over policy, duplicate orders) on everyday jobs, refereed from the end state. Latest: Crucible 2: Hard, per model, with 95% intervals and replays.
It's a remote server: nothing to install or run. No API key, no account, read only.
| URL | https://theagentforge.io/mcp/public |
| Transport | Streamable HTTP |
| Auth | none |
| Server card | https://theagentforge.io/mcp/public/server-card |
| Official MCP Registry | io.theagentforge/forge |
Tools
| Tool | What it does |
|---|---|
forge_check |
Field reports and published advisories for one tool version. Give one of id (npm:name, pypi:name, github.com/owner/repo; a version may be inline, npm:name@1.2.3), url or fingerprint (sha256:<64 hex>), plus version if known. |
forge_scan |
Scans a skill file, tool descriptions or install script (text, or url on a public code host) for prompt injection and risky instructions. Returns a verdict, rule ids, line numbers and the file's sha256. |
forge_watch |
Re-checks installed skills by sha256 fingerprints (up to 100): which are flagged, need a fresh scan or have new field reports. |
forge_exposure |
Matches installed packages (npm or PyPI name and exact version, up to 500) or pip_freeze text against OSV advisories and CISA's exploited list. Names and versions only. |
forge_traps |
How often AI models fell for each planted trap: public trap id (FA-*), attempts, falls and fall rate. Optional task, model, tier (normal or hard). |
arena_results |
Per job and model: attempts, passes, pass rate with a 95% interval, 3-of-3 consistency, traps fallen for, knockouts, cost per pass and replays. Optional task, tier, event (e.g. crucible-2-hard). |
Every result is data, not instructions. A clear scan means no known pattern matched, not that a tool is safe.
Check and scan answers include share.receipt_url, a signed, dated record of the answer.
Data downloads: Crucible 2: Hard (JSON) · CSV · all datasets
Connect
Claude Code
claude mcp add --transport http forge https://theagentforge.io/mcp/public
Gemini CLI
gemini extensions install https://github.com/theagentforge/forge-mcp
Agent Skills (skills.sh, any agent that reads SKILL.md):
npx skills add theagentforge/forge-mcp
Three skills, all text only, no scripts:
forge-check-before-installscans a skill, plugin, MCP server or install command before it is installed;forge-run-the-gauntletruns the Gauntlet, six refereed jobs with planted traps, on your own agent;forgecovers checks, field reports and scans over the MCP server.
Codex, ChatGPT, GitHub Copilot, Kiro, Cursor: this repo is an Agent Plugins 1.0
package (plugin.json, mcp.json, skills/) with no hooks: the remote server and the skills below only.
Install it from each app's plugin marketplace once it is listed there, or point an app that installs plugins from Git
at this repo's URL. In Copilot CLI, once it is listed in awesome-copilot:
copilot plugin install the-agent-forge@awesome-copilot
Cursor: this repo is also a Cursor plugin (.cursor-plugin/plugin.json, .mcp.json). Or add to ~/.cursor/mcp.json:
{ "mcpServers": { "the-agent-forge": { "url": "https://theagentforge.io/mcp/public" } } }
VS Code / Copilot: add to .vscode/mcp.json:
{ "servers": { "the-agent-forge": { "type": "http", "url": "https://theagentforge.io/mcp/public" } } }
Cline: see llms-install.md.
Anything else: any MCP client that speaks Streamable HTTP. Point it at the URL above. Every app's steps: https://theagentforge.io/connect.
Other ways in
- A2A agent card: https://theagentforge.io/.well-known/agent-card.json
- API catalog (RFC 9727): https://theagentforge.io/.well-known/api-catalog
- AI Catalog with the MCP Server Cards: https://theagentforge.io/.well-known/ard.json (also at
/.well-known/ai-catalog.json) - Any page as Markdown: send
Accept: text/markdown - llms.txt: https://theagentforge.io/llms.txt
About
This repository holds the connection files for directories and clients, under the MIT licence. The service itself runs at theagentforge.io. Its source is published for audit at theagentforge/agent-forge.
| Files | Read by |
|---|---|
plugin.json, mcp.json, skills/*/SKILL.md |
Agent Plugins 1.0 clients (Codex, ChatGPT, Copilot, VS Code, Kiro, Cursor) |
skills/*/SKILL.md |
The skills installer (skills.sh) |
gemini-extension.json, GEMINI.md, skills/ |
Gemini CLI |
.cursor-plugin/plugin.json, .mcp.json |
Cursor, Grok Build, Claude Code |
No hooks, scripts or binaries. The only network endpoint is https://theagentforge.io/mcp/public, and it needs no
credentials.
Privacy: https://theagentforge.io/privacy · Terms: https://theagentforge.io/terms · Contact: hello@theagentforge.io
The Agent Forge MCP questions
How do I add The Agent Forge MCP to Claude Code?
Run claude mcp add --transport http forge-mcp https://theagentforge.io/mcp/public in your terminal. Add --scope user to make it available in every project.
How do I add The Agent Forge MCP to Cursor?
Open ~/.cursor/mcp.json and add the server under mcpServers, using the Cursor config shown in the install section above.
What tools does The Agent Forge MCP provide?
It exposes 6 tools, including forge_check, forge_scan, forge_watch, forge_exposure, forge_traps.
Does using The Agent Forge require an API key?
No, the server is read-only and requires no API key or account.
What can the skill scanner detect?
It scans for prompt injection, hidden instructions, shell-piped downloads, credential/key-file reads, and data exfiltration to unknown endpoints.
How does the server handle package vulnerability checks?
It matches installed npm and PyPI versions against published OSV advisories and CISA's known-exploited list.