VulnForge MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| vulnforge_audit_webB | Audits web HTTP response headers, cookie flags, and DOM client code for security misconfigurations, XSS sinks, weak CSP, HSTS absences, and tabnabbing. |
| vulnforge_audit_apiB | Audits REST/GraphQL endpoint handlers for Broken Object Level Auth (BOLA/IDOR), unauthenticated mutating endpoints, missing rate limits, and mass assignment risks. |
| vulnforge_audit_dbB | Audits SQL/NoSQL query snippets, ORM commands, and database connection strings for injection vulnerabilities, raw interpolations, and credential leaks. |
| vulnforge_audit_graphqlB | Audits GraphQL schemas, resolvers, queries, and server configurations for introspection exposure, cyclic query depth DoS, query batching attacks, field complexity bypass, and sensitive field leakage without authorization. |
| vulnforge_inject_patchB | Synthesizes and automatically injects defensive patches into vulnerable code, converting raw queries to parameterized statements, adding headers, sanitizing DOM sinks, and optionally verifying via canary sandbox. |
| vulnforge_verify_patchB | Executes isolated Node.js vm canary sandbox verification of defensive patches against dual attack and benign functional payloads to prove exploit mitigation with zero regressions. |
| vulnforge_audit_cicdA | Audits GitHub Actions (.github/workflows/*.yml) and GitLab CI (.gitlab-ci.yml) workflows for dangerous triggers (pull_request_target with untrusted checkout), missing or overprivileged permissions (write-all), hardcoded secrets/tokens, unpinned third-party actions, context expression script injections, and insecure OIDC configurations. |
| vulnforge_audit_containerA | Audits Dockerfiles and Kubernetes manifests (Pods, Deployments, RBAC Roles, ClusterRoles) for root container execution, missing multi-stage builds, unpinned mutable images, hardcoded secrets, dangerous capabilities, host namespace sharing, and unconstrained RBAC wildcards. |
| vulnforge_audit_ebpfC | Audits eBPF C programs, Falco rules, Tetragon TracingPolicies, and Seccomp profiles for kernel verifier safety, unbounded loops, unchecked map lookup pointers, dangerous helpers (bpf_probe_write_user), direct packet access OOBs, missing container shell spawn rules, and permissive default actions. |
| vulnforge_fuzz_apiB | Fuzzes REST/API route handlers and schemas using dynamic feedback-guided mutations (boundary values, type mismatches, prototype pollution probes, oversized payloads) to uncover unhandled crashes and input validation gaps, synthesizing defensive Zod/sanitizer validation patches. |
| vulnforge_audit_iamA | Audits Cloud-Native IAM policies and Zero Trust access rules (AWS IAM JSON, Terraform HCL, GCP IAM bindings, Azure RBAC) for overprivileged wildcards, privilege escalation paths (iam:PassRole), insecure public trust relationships, confused deputy risks, and missing Zero Trust TLS/MFA controls. |
| vulnforge_audit_vaultA | Audits secret rotation policies, cryptographic key lifecycles, and vault configurations across HashiCorp Vault HCL/JSON, AWS Secrets Manager, Azure Key Vault, GCP Secret Manager, and Kubernetes KMS EncryptionConfig for stale rotation intervals, weak key derivation/ciphers, excessive lease TTLs, unencrypted backends, missing audit logging, wildcard secret policies, single unseal threshold risks, and plaintext secret exposures. |
| vulnforge_audit_sbomA | Audits Software Bill of Materials (SBOM) manifests (CycloneDX JSON/XML, SPDX JSON/Tag-Value), package manifests (package.json, package-lock.json, Cargo.toml, requirements.txt), and registry configs (.npmrc) for dependency confusion (CWE-427), lookalike typosquatting (CWE-829), compromised packages (event-stream, flatmap-stream), unpinned wildcards (CWE-1104), dangerous install lifecycle scripts, missing integrity hashes, and NTIA minimum elements compliance (EO 14028). |
| vulnforge_audit_cosignC | Audits Artifact Cosign signatures, Sigstore keyless attestation proofs, Rekor transparency log SET inclusion, in-toto SLSA Build Level 3 provenance, and Kyverno/Gatekeeper container admission verification policies against 8 core supply chain vulnerability vectors (CWE-347, CWE-295, CWE-353, CWE-829, CWE-298, CWE-327, CWE-1104, CWE-284). |
| vulnforge_audit_wasm_ebpfB | Audits WebAssembly runtime configurations (WasmEdge, Wasmtime, Envoy proxy-wasm filters), eBPF-to-Wasm hostcall bridges, and distributed host telemetry streams against 8 critical vectors (CWE-119 memory bounds, CWE-284 WASI root preopens, CWE-400 fuel/CPU DoS, CWE-345 unpinned digest, CWE-778 dropped telemetry, CWE-209 credential leaks, CWE-276 shared memory permissions, CWE-693 unconfined host privileges), with causal multi-host attack chain correlation and automated secret scrubbing. |
| vulnforge_audit_microsegmentationB | Audits distributed eBPF microsegmentation policies (CiliumNetworkPolicy, Kubernetes NetworkPolicy), verifies Zero-Trust default-deny postures, cloud metadata egress restrictions (169.254.169.254/32), L7 protocol visibility, and auto-synthesizes hardened policies from live kernel flow telemetry. |
| vulnforge_audit_zkp_gatewayA | Audits Zero-Knowledge Proof (ZKP) verification gates, Groth16/Plonk arithmetic circuits, W3C Verifiable Credential revocation registries, and DID authentication gateways across 8 critical cryptographic and gateway vectors (CWE-682 underconstrained signals, CWE-20 unvalidated public inputs, CWE-294 nullifier replay & double-spending, CWE-284 mock verifier bypasses, CWE-345 unpinned verification keys, CWE-295 missing revocation registries, CWE-613 expired proof lifetimes, CWE-327 weak Fiat-Shamir transcripts) with automated policy synthesis and canary sandbox verification. |
| vulnforge_audit_model_supply_chainA | Audits AI model supply chains, Safetensors tensor headers, PyTorch pickle opcodes (CWE-502), and model hub provenance configs across 8 critical vectors (CWE-502, CWE-345, CWE-20, CWE-494, CWE-295, CWE-829, CWE-200, CWE-778) with automated patch synthesis and canary sandbox verification. |
| vulnforge_audit_intotoC | Audits In-toto supply chain metadata, DSSE envelopes, cryptographic provenance attestations, SLSA predicate policies, and Sigstore Rekor transparency log proofs across 8 critical vulnerability vectors (CWE-347, CWE-353, CWE-829, CWE-778, CWE-284, CWE-20, CWE-613, CWE-200) with automated patch synthesis and canary verification. |
| vulnforge_audit_kms_hsmC | Audits Multi-Tenant KMS Key Derivation, Cloud KMS Lifecycles, and Hardware Security Module (HSM) Remote Attestation across 8 critical vulnerability vectors (CWE-320, CWE-284, CWE-295, CWE-732, CWE-361, CWE-327, CWE-522, CWE-778) with automated patch synthesis and canary verification. |
| vulnforge_fuzz_wasm_runtimeC | Audits WebAssembly execution engines (Wasmtime, Wasmer, WasmEdge), WASI capability preopens, and differential mutation fuzzing specs across 8 critical vulnerability vectors (CWE-119, CWE-284, CWE-400, CWE-770, CWE-698, CWE-362, CWE-459, CWE-778) with automated policy synthesis and canary sandbox verification. |
| vulnforge_audit_p2p_gossipC | Audits Decentralized P2P Gossip Protocols, Policy Revocation Manifests, Byzantine Quorum Consensus, and Distributed Transparency Log Sync across 8 critical vulnerability vectors (CWE-347, CWE-284, CWE-294, CWE-362, CWE-400, CWE-353, CWE-200, CWE-754) with automated policy synthesis and canary sandbox verification. |
| vulnforge_audit_pqcC | Audits Post-Quantum Hybrid Key Encapsulation (ML-KEM / Kyber), FIPS 203 Cryptographic Migration, TLS 1.3 / SSH KEX specs, and X.509 hybrid certificates across 8 critical vulnerability vectors (CWE-326, CWE-327, CWE-20, CWE-459, CWE-754, CWE-319) with automated policy synthesis and canary sandbox verification. |
| vulnforge_full_scanA | Runs an end-to-end multi-layer vulnerability audit across Web, API, Database, GraphQL, CI/CD, Container/Kubernetes, and eBPF/Runtime Security layers, generating unified findings and ready-to-apply patches. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 24 tools
Most tools target distinct security domains, but several clusters overlap heavily: audit_sbom, audit_cosign, audit_intoto, audit_model_supply_chain, and audit_p2p_gossip all cover supply-chain/attestation concerns, while audit_ebpf, audit_wasm_ebpf, audit_microsegmentation, and fuzz_wasm_runtime overlap runtime/kernel surfaces. Descriptions do help differentiate them, but an agent could still misselect when asked for a broad layer such as 'supply chain audit'. The presence of full_scan alongside granular audits adds another selection ambiguity.
Almost all tools use a consistent vulnforge_ prefix with snake_case verb/noun patterns such as audit_*, fuzz_*, inject_patch, and verify_patch. The only minor deviations are full_scan lacking the audit/fuzz verb style and some names combining multiple nouns, but the convention remains predictable overall.
Twenty-four tools is heavy for a single server and likely reflects a very broad security audit suite rather than a tightly scoped toolset. Most tools do earn their place by covering distinct security domains, but the count is borderline excessive and increases selection burden.
The surface covers a wide audit lifecycle: domain-specific audits, fuzzing, patch injection, patch verification, and an end-to-end full scan. Minor gaps exist around reporting/export, triage workflows, and general result management, but agents can work around these for most audit workflows.