Security & Testing
vulntel.com
A hosted MCP server that provides AI agents with source-grounded vulnerability intelligence from NVD, CISA KEV, EPSS, OSV/GHSA, and SSVC for exploitation-first answers, by-mechanism search, and CVE fact-checking.
ENDPOINT 1
https://mcp.rozetyp.com/mcp
MCP server metadata
- Name
- vulntel
- Version
- 1.2.0
Vulnerability intelligence for authorized security research and bug bounty triage. Returns ranked, source-grounded data fused from NVD + CISA KEV + FIRST EPSS + OSV/GHSA + CISA Vulnrichment (SSVC: exploitation/automatable/technical-impact), with live public-exploit (PoC) signal, plus Debian / Ubuntu / Alpine security-tracker fixed builds (distros backport fixes without changing the upstream version: pass server banners VERBATIM, e.g. 'OpenSSH_8.9p1 Ubuntu-3ubuntu0.10', 'Apache/2.4.62 (Debian)', and each CVE gets `distro_fix` next to the upstream `affected`). Use new_in_scope to find assets that just entered a bug-bounty program's scope (fresh surface, re-adds and renames filtered out), hunt_plan to turn a recon'd stack into a ranked dig-order (+ recurring loci), check_technology for one product's version-specific CVEs, search_vulns to find CVEs by concept/keyword or by a seed CVE's mechanism (it fuses keyword + semantic search, and also serves the recent high-risk window via days=/product=), verify_cve_claim to fact-check a CVE assertion (exists? affects this version? severity? exploited?) and catch hallucinations, enrich_cve to dig into one finding, and search_public_code to find public repos that copied a given code string (raw hits -- YOU read them to judge patched-vs-vulnerable; best for learning how a bug was fixed in the wild or spotting unpatched vendored copies, not for finding exploitable targets). The server lays out facts; you do the reasoning. Output is for triage and prioritization, not exploitation. vulntel is free and run by one person (rozetyp). Feedback and corrections: https://github.com/rozetyp/vulntel-mcp/issues. What it keeps: https://vulntel.com/data-handling.
Known tools 13
enrich_cveFull intelligence for a single CVE: CVSS, KEV status, EPSS, CISA SSVC (exploitation/automatable/technical_impact), P1-P4 priority with rationale, affected products/versions, Debian/Ubuntu/Alpine fixed builds per release (`distro_fixes`), and (live) public-exploit signal.
Inferred read-onlycheck_technologyFind vulnerabilities affecting a product across NVD (CPE) and OSV (package ecosystems), de-duplicated and ranked by priority/EPSS/CVSS.
Inferred read-onlyobserveScan a host's exposed stack and join it to the CVE corpus — the recon workhorse.
Inferred read-onlysearch_vulnsSearch the CVE corpus — one entry point, three retrieval modes picked from your arguments: - query= a concept / technique / weakness (e.
Inferred read-onlyfind_attack_approachesFind how bugs *like this* were actually FOUND — retrieve real disclosed bug-bounty report mechanic-cards whose approach is closest (by mechanism) to your query or a seed CVE.
Inferred read-onlyfind_continuationsWhen you're STUCK mid-hunt: describe your POSITION — the attacker state/capabilities you hold, stripped of product names — and get the CONTINUATIONS that worked from positions like it, matched against the full disclosed-report BODIES (the concrete step/payload/precondition), not just the distilled card.
Inferred read-onlyprogram_outcome_priorAt the START of an engagement on a bug-bounty PROGRAM, get a grounded prior over what bug classes have historically LANDED there — to decide what to look at first and what to skip.
Inferred read-onlynew_in_scopeFind FRESH attack surface: bounty-eligible assets that entered a public bug-bounty program's scope in the last `days` (HackerOne, Bugcrowd, Intigriti, YesWeHack, mirrored daily), the least-hunted surface there is.
Inferred read-onlyverify_cve_claimFact-check a claim about a CVE against the corpus — catch hallucinations before they ship.
Inferred read-onlycorpus_statsCorpus size and data freshness (CVE / KEV / EPSS counts and latest timestamps).
Inferred read-onlysearch_public_codeSearch public repositories for an exact code string; return WHERE it appears.
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.vulntel]
url = "https://mcp.rozetyp.com/mcp"
enabled = true
Claude Code
.mcp.json
{
"mcpServers": {
"vulntel": {
"type": "http",
"url": "https://mcp.rozetyp.com/mcp"
}
}
}
Claude Desktop
Settings → Connectors → Add custom connector
Name: vulntel
Remote MCP URL: https://mcp.rozetyp.com/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": {
"vulntel": {
"url": "https://mcp.rozetyp.com/mcp"
}
}
}
Visual Studio Code
.vscode/mcp.json
Add to Visual Studio Code{
"servers": {
"vulntel": {
"type": "http",
"url": "https://mcp.rozetyp.com/mcp"
}
}
}
Generic MCP
Client-specific MCP configuration
{
"name": "vulntel",
"transport": "streamable-http",
"url": "https://mcp.rozetyp.com/mcp"
}
MCP Inspector
Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.
ENDPOINT 2
https://mcp.vulntel.com/mcp
MCP server metadata
- Name
- vulntel
- Version
- 1.30.0
Vulnerability intelligence for authorized security research and bug bounty triage. Returns ranked, source-grounded data fused from NVD + CISA KEV + FIRST EPSS + OSV/GHSA + CISA Vulnrichment (SSVC: exploitation/automatable/technical-impact), with live public-exploit (PoC) signal, plus Debian / Ubuntu / Alpine security-tracker fixed builds (distros backport fixes without changing the upstream version: pass server banners VERBATIM, e.g. 'OpenSSH_8.9p1 Ubuntu-3ubuntu0.10', 'Apache/2.4.62 (Debian)', and each CVE gets `distro_fix` next to the upstream `affected`). Use new_in_scope to find assets that just entered a bug-bounty program's scope (fresh, least-hunted surface, re-adds and renames filtered out), hunt_plan to turn a recon'd stack into a ranked dig-order (+ recurring loci), check_technology for one product's version-specific CVEs, search_vulns to find CVEs by concept/keyword or by a seed CVE's mechanism (it fuses keyword + semantic search, and also serves the recent high-risk window via days=/product=), verify_cve_claim to fact-check a CVE assertion (exists? affects this version? severity? exploited?) and catch hallucinations, enrich_cve to dig into one finding, and search_public_code to find public repos that copied a given code string (raw hits -- YOU read them to judge patched-vs-vulnerable; best for learning how a bug was fixed in the wild or spotting unpatched vendored copies, not for finding exploitable targets). The server lays out facts; you do the reasoning. Output is for triage and prioritization, not exploitation.
Known tools 13
enrich_cveFull intelligence for a single CVE: CVSS, KEV status, EPSS, CISA SSVC (exploitation/automatable/technical_impact), P1-P4 priority with rationale, affected products/versions, Debian/Ubuntu/Alpine fixed builds per release (`distro_fixes`), and (live) public-exploit signal.
Inferred read-onlycheck_technologyFind vulnerabilities affecting a product across NVD (CPE) and OSV (package ecosystems), de-duplicated and ranked by priority/EPSS/CVSS.
Inferred read-onlyobserveScan a host's exposed stack and join it to the CVE corpus — the recon workhorse.
Inferred read-onlysearch_vulnsSearch the CVE corpus — one entry point, three retrieval modes picked from your arguments: - query= a concept / technique / weakness (e.
Inferred read-onlyfind_attack_approachesFind how bugs *like this* were actually FOUND — retrieve real disclosed bug-bounty report mechanic-cards whose approach is closest (by mechanism) to your query or a seed CVE.
Inferred read-onlyfind_continuationsWhen you're STUCK mid-hunt: describe your POSITION — the attacker state/capabilities you hold, stripped of product names — and get the CONTINUATIONS that worked from positions like it, matched against the full disclosed-report BODIES (the concrete step/payload/precondition), not just the distilled card.
Inferred read-onlyprogram_outcome_priorAt the START of an engagement on a bug-bounty PROGRAM, get a grounded prior over what bug classes have historically LANDED there — to decide what to look at first and what to skip.
Inferred read-onlynew_in_scopeFind FRESH attack surface: bounty-eligible assets that entered a public bug-bounty program's scope in the last `days` (HackerOne, Bugcrowd, Intigriti, YesWeHack, mirrored daily), the least-hunted surface there is.
Inferred read-onlyverify_cve_claimFact-check a claim about a CVE against the corpus — catch hallucinations before they ship.
Inferred read-onlycorpus_statsCorpus size and data freshness (CVE / KEV / EPSS counts and latest timestamps).
Inferred read-onlysearch_public_codeSearch public repositories for an exact code string; return WHERE it appears.
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.vulntel]
url = "https://mcp.vulntel.com/mcp"
enabled = true
Claude Code
.mcp.json
{
"mcpServers": {
"vulntel": {
"type": "http",
"url": "https://mcp.vulntel.com/mcp"
}
}
}
Claude Desktop
Settings → Connectors → Add custom connector
Name: vulntel
Remote MCP URL: https://mcp.vulntel.com/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": {
"vulntel": {
"url": "https://mcp.vulntel.com/mcp"
}
}
}
Visual Studio Code
.vscode/mcp.json
Add to Visual Studio Code{
"servers": {
"vulntel": {
"type": "http",
"url": "https://mcp.vulntel.com/mcp"
}
}
}
Generic MCP
Client-specific MCP configuration
{
"name": "vulntel",
"transport": "streamable-http",
"url": "https://mcp.vulntel.com/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.