// docs / mcp
MCP सेवकः
Plug FixWeb into Claude Desktop, Cursor, वा any client that speaks तत् Model Context Protocol. Your AI agent gets typed access प्रति तव scans, findings, च तत् same templated fix prompts that power तत् dashboard's प्रतिलिपि fix prompt button.
Mint an API token
Visit /लेखा/api-tokens च create a token named, e.g., claude-desktop. प्रतिलिपि तत् plaintext value — it's shown once.
Tokens सन्ति bearer credentials: anyone सह तत् string शक्नोति read तव scans च start नूतनम् ones. Store it like a password.
Point तव MCP client at /api/mcp
Claude Desktop / Cursor / Continue / Zed:
{
"mcpServers": {
"fixweb": {
"transport": "streamable-http",
"url": "https://fixweb.app/api/mcp",
"headers": {
"Authorization": "Bearer fxw_YOUR_TOKEN_HERE"
}
}
}
}Restart तत् client. The fixweb server should appear in its MCP सेवकः list.
Try it out
Ask तव agent things like:
- “List my last 10 FixWeb scans.”
- “Show me तत् critical findings on तत् most recent scan.”
- “Start a passive scan against
https://staging.example.com.” - “For each high-severity finding on scan X, write a fix.”
- “Are there any open live-threat alerts on my domains?”
- Type
/fixweb-fixसह a finding id प्रति drop तत् templated remediation prompt straight into तत् chat.
उपकरणानि
- list_scansread
- Returns up प्रति 100 most-recent scans सह status + finding counts. Args: limit?: 1..100.
- get_scanread
- Scan envelope + per-category severity summary by default. Set include_findings=true कृते तत् full वृत्तान्तः (large कृते noisy scans — prefer list_findings + filters). Args: scan_id (uuid), include_findings?: boolean.
- list_findingsread
- Paginated findings across all तव scans. Args: severity?: list, check_id?, since? (ISO 8601), limit?: 1..200.
- start_scanwrite
- Enqueues a passive scan. Returns an id with status queued; poll get_scan to await completion. Owner-depth mode is gated behind on-site attestation and not exposed via MCP. Args: target (URL or hostname).
- list_alertsread
- जीवन्त-भयसूचनाः (CT log भेदाः, DNS परिवर्तनानि, threat intel सूचयः)। केवलम् Unlimited योजनायाम् उपलब्धम्; Hobby Pro योजनाः रिक्त-सूचीं प्रत्यर्पयन्ति। Args: domain_id?, active_only?, limit?: 1..200.
- get_alertread
- Single alert सह full payload (DNS diff, नूतनम् certs, listing detail). Args: alert_id (uuid).
- dismiss_alertwrite · idempotent
- Mark an alert dismissed. Idempotent — re-dismissing अस्ति a न-op. Args: alert_id (uuid).
साधनानि
साधनानि let तव client attach FixWeb दत्तांशः into तत् conversation directly, instead of तत् agent re-fetching it on प्रत्येकम् turn. In Claude Desktop, click तत् @ menu → fixweb.
- fixweb://scan/{scan_id}/reportjson
- Full FixWeb scan वृत्तान्तः including प्रत्येकम् check च प्रत्येकम् finding.
- fixweb://finding/{finding_id}json
- A single finding (severity, title, description, evidence, remediation, CWE).
Slash commands
- /fixweb-fixprompt
- Renders a templated remediation prompt कृते a finding. Detects तत् codebase framework तः तत् scan's tech-fingerprint च injects framework-specific advice यदा available; falls back प्रति a generic recipe otherwise. Args: finding_id (uuid). No Claude API call — templated server-side.
→ Quotas, RLS, च severity gating apply identically प्रति MCP च REST calls.
