Power Strip Security MCP
An MCP server that provides prompt-injection shielding for AI agents. It cleans web pages by removing hidden text and invisible characters, scans content for instruction-override phrases, and audits MCP servers and agent skills for tool poisoning or malicious scripts before connection.
Endpoint and access
No sign-in to connect and list tools. Individual tools may still need a vendor account, credentials, or permissions; check the documented prerequisites.
https://api.powerstripsecurity.workers.dev/mcpWhat you can do with it
- Read a URL and get clean markdown with hidden injections removed
- Scan HTML or text content for prompt injection risks
- Audit an MCP server URL for tool poisoning before connecting
- Audit agent skill files for hardcoded keys or exfiltration
Source and maintenance
Protocol and capabilities
- Server
- power-strip-security 0.0.39
- MCP protocol
- 2025-06-18
- Tools
- 4 advertised
- Resources
- Not advertised
- Resource templates
- Not advertised
- Prompts
- Not advertised
{
"tools": {}
}Server-provided instructions
Untrusted server output shown for inspection. It is not Koskamo guidance and is excluded from knowledge search.
Tools
4 advertised
read_url — Read a web page or PDF safely
Fetch a public web page or PDF with hidden text and exfiltration links removed, plus a prompt-injection report. Use it instead of fetching untrusted URLs yourself; for content you already have, use check_prompt_injection. Returns risk (0-1), level, findings, content and wrapped_text: put wrapped_text in your prompt, not the raw page. Costs US$0.005 in USDC on Base (x402); unreadable pages (robots.txt, login, CAPTCHA, JavaScript-only) return a reason and are not charged.
{
"type": "object",
"required": [
"url"
],
"properties": {
"url": {
"type": "string",
"format": "uri",
"description": "Public http(s) URL of an HTML page, text file or PDF (up to 5 redirects)"
},
"format": {
"enum": [
"markdown",
"text"
],
"type": "string",
"description": "Output format of content: markdown (default) keeps links and headings; text is plain text without links"
}
}
}{
"readOnlyHint": true,
"openWorldHint": true,
"destructiveHint": false
}check_prompt_injection — Scan content for prompt injection
Scan HTML or text you already have (an email, a document, a tool result, a page fetched elsewhere) for prompt injection before it reaches your prompt. To fetch a URL, use read_url instead; for MCP tool definitions or skill files, use audit_mcp_server or audit_agent_skill. Returns risk (0-1), level (low/medium/high), findings (type, snippet, position, confidence), clean_text and wrapped_text, which marks the rest as untrusted data. Costs US$0.002 in USDC on Base (x402). Max 256 KB.
{
"type": "object",
"required": [
"content"
],
"properties": {
"url": {
"type": "string",
"description": "Where the content came from (optional); used to resolve relative links in HTML"
},
"format": {
"enum": [
"html",
"text"
],
"type": "string",
"description": "How to parse content: html (default) also strips hidden elements, comments and styles; use text for plain text such as an email body"
},
"content": {
"type": "string",
"description": "The HTML or text to scan (max 256 KB)"
},
"return_text": {
"type": "boolean",
"description": "Include clean_text and wrapped_text (default true); false returns only the report, which is smaller"
}
}
}{
"readOnlyHint": true,
"openWorldHint": true,
"destructiveHint": false
}audit_mcp_server — Audit an MCP server for tool poisoning
Check an MCP server for tool poisoning before connecting: hidden instructions, sensitive-file access, exfiltration parameters, cross-tool shadowing, rug pulls. For skills or AGENTS.md use audit_agent_skill. Give url (tools are listed read-only, never called) or the pasted tools, not both. Returns level, findings (type, severity, location, snippet) and per-tool fingerprints to send back as pinned. Costs US$0.01 in USDC on Base (x402); servers that cannot be listed are not charged.
{
"type": "object",
"properties": {
"url": {
"type": "string",
"format": "uri",
"description": "Streamable HTTP endpoint of a public MCP server. Omit when pasting tools"
},
"tools": {
"type": "array",
"items": {
"type": "object"
},
"description": "Instead of url: the tools array of a tools/list result, pasted as is"
},
"pinned": {
"type": "object",
"description": "Optional map of tool name to the fingerprint returned by a previous audit; any change is reported as tool-changed (high)",
"additionalProperties": {
"type": "string"
}
},
"instructions": {
"type": "string",
"description": "Only with tools: the server's instructions from its initialize result"
}
}
}{
"readOnlyHint": true,
"openWorldHint": true,
"destructiveHint": false
}audit_agent_skill — Audit an agent skill or AGENTS.md
Check an agent skill (SKILL.md with its scripts and references) or AGENTS.md, CLAUDE.md, .cursorrules before loading it: prompt injection, curl | bash, encoded commands, credential theft, exfiltration, hardcoded keys, persistence. For MCP servers use audit_mcp_server. Give files or one url, not both. Returns level, findings (type, severity, file, line, snippet) and per-file fingerprints to send back as pinned. Costs US$0.005 in USDC on Base (x402); nothing is executed.
{
"type": "object",
"properties": {
"url": {
"type": "string",
"format": "uri",
"description": "Instead of files: public URL of one raw text file, e.g. a SKILL.md on raw.githubusercontent.com (github.com/.../blob/... links are converted)"
},
"files": {
"type": "array",
"items": {
"type": "object",
"required": [
"path",
"content"
],
"properties": {
"path": {
"type": "string"
},
"content": {
"type": "string"
}
}
},
"description": "The skill files with their relative paths, e.g. [{path: 'SKILL.md', content: '...'}, {path: 'scripts/setup.sh', content: '...'}] (max 50 files, 256 KB in total). Omit when using url"
},
"pinned": {
"type": "object",
"description": "Optional map of file path to the fingerprint returned by a previous audit; changed files are reported",
"additionalProperties": {
"type": "string"
}
}
}
}{
"readOnlyHint": true,
"openWorldHint": true,
"destructiveHint": false
}Prompts
Not advertised
Resources
Not advertised
30-day availability
Aggregated from read-only remote checks; this is not a tool-execution or security guarantee.
- Successful checks
- 1 of 1 (100%)
- Average latency
- 987 ms
- 95th percentile latency
- 987 ms
Maintenance history
Public check and source-refresh events. Editorial and ownership records remain private.
- · probe · passed · 987 ms · metadata changed · 4 tool(s)
Connect from your client
Each snippet registers the same endpoint; pick the client you use.
Claude Code
claude mcp add --transport http power-strip-security-mcp https://api.powerstripsecurity.workers.dev/mcpCodex
codex mcp add power-strip-security-mcp --url https://api.powerstripsecurity.workers.dev/mcpCursor
{
"mcpServers": {
"power-strip-security-mcp": {
"url": "https://api.powerstripsecurity.workers.dev/mcp"
}
}
}Cline
{
"mcpServers": {
"power-strip-security-mcp": {
"type": "streamableHttp",
"url": "https://api.powerstripsecurity.workers.dev/mcp",
"disabled": false
}
}
}VS Code
{
"servers": {
"power-strip-security-mcp": {
"type": "http",
"url": "https://api.powerstripsecurity.workers.dev/mcp"
}
}
}Gemini CLI
gemini mcp add --transport http power-strip-security-mcp https://api.powerstripsecurity.workers.dev/mcpClaude Desktop
{
"mcpServers": {
"power-strip-security-mcp": {
"url": "https://api.powerstripsecurity.workers.dev/mcp"
}
}
}Windsurf
{
"mcpServers": {
"power-strip-security-mcp": {
"serverUrl": "https://api.powerstripsecurity.workers.dev/mcp"
}
}
}Let your agent find its tools
Find servers like Power Strip Security MCP from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.