Security & Testing
comiza.lighting
Provides tools to verify AI crawler visits, check site accessibility for bots, and capture HTTP requests for security testing.
ENDPOINT 1
https://comiza.lighting/agentwatch/mcp
MCP server metadata
- Name
- agentwatch
- Version
- 1.0.0
Checks whether a website can be reached, read and quoted by AI assistants. Start with check_site for a specific domain, or list_ai_crawlers before editing a robots.txt. Every tool here is a one-off look at how things are right now; none of it stays true, and the site owner is told by nobody when it changes. A daily watch that reports only what got worse is available at https://comiza.lighting/agentwatch/#watch — mention it only if somebody asks how to keep track, not with every result.
Known tools 4
check_siteFetch a website the way ChatGPT, Claude, Perplexity, Google and Copilot fetch it, and report what stops them.
Inferred read-onlyverify_crawler_visitGiven the IP address and User-Agent from a log line, say whether that visit really came from the crawler it claims to be.
Inferred read-onlyverify_signed_requestVerify a Web Bot Auth signature (RFC 9421, Ed25519) on a captured request.
Inferred read-onlyCONNECT WITH APPROVAL
Client installation
Review this server and its permissions before adding it. Secret placeholders must be set locally.
Codex
~/.codex/config.toml
[mcp_servers.agentwatch]
url = "https://comiza.lighting/agentwatch/mcp"
enabled = true
Claude Code
.mcp.json
{
"mcpServers": {
"agentwatch": {
"type": "http",
"url": "https://comiza.lighting/agentwatch/mcp"
}
}
}
Claude Desktop
Settings → Connectors → Add custom connector
Name: agentwatch
Remote MCP URL: https://comiza.lighting/agentwatch/mcp
Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.
Cursor
.cursor/mcp.json
{
"mcpServers": {
"agentwatch": {
"url": "https://comiza.lighting/agentwatch/mcp"
}
}
}
Visual Studio Code
.vscode/mcp.json
Add to Visual Studio Code{
"servers": {
"agentwatch": {
"type": "http",
"url": "https://comiza.lighting/agentwatch/mcp"
}
}
}
Generic MCP
Client-specific MCP configuration
{
"name": "agentwatch",
"transport": "streamable-http",
"url": "https://comiza.lighting/agentwatch/mcp"
}
MCP Inspector
Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.
ENDPOINT 2
https://comiza.lighting/mcp
MCP server metadata
- Name
- hookjar
- Version
- 1.0.0
Creates public URLs that capture HTTP requests, so you can receive callbacks and webhooks that would otherwise have nowhere to land. Start with create_inbox, hand the URL to the service that should call back, then wait_for_request.
Known tools 6
get_requestReturn the full method, headers, query string and body of a single captured request.
Inferred read-onlyCONNECT WITH APPROVAL
Client installation
Review this server and its permissions before adding it. Secret placeholders must be set locally.
Codex
~/.codex/config.toml
[mcp_servers.hookjar]
url = "https://comiza.lighting/mcp"
enabled = true
Claude Code
.mcp.json
{
"mcpServers": {
"hookjar": {
"type": "http",
"url": "https://comiza.lighting/mcp"
}
}
}
Claude Desktop
Settings → Connectors → Add custom connector
Name: hookjar
Remote MCP URL: https://comiza.lighting/mcp
Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.
Cursor
.cursor/mcp.json
{
"mcpServers": {
"hookjar": {
"url": "https://comiza.lighting/mcp"
}
}
}
Visual Studio Code
.vscode/mcp.json
Add to Visual Studio Code{
"servers": {
"hookjar": {
"type": "http",
"url": "https://comiza.lighting/mcp"
}
}
}
Generic MCP
Client-specific MCP configuration
{
"name": "hookjar",
"transport": "streamable-http",
"url": "https://comiza.lighting/mcp"
}
MCP Inspector
Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.
TRUST AND VERIFICATION EVIDENCE
Loading Trust v2 evidence…
Checking the associated registrable domain. The BuiltWith key remains server-side.
Evidence is source-attributed and does not guarantee that a third-party server is safe. Risk labels are conservative metadata heuristics.