OWASP MCP Top 10 — Compliance Assessment
2025 · 2026-04-23 registry entry
Executive summary
✗ Non-compliant
Assessment of dodopayments-mcp against OWASP MCP Top 10: overall status non compliant. Of 10 controls, 2 met, 5 unmet, 0 partial, 3 not assessed, 0 not applicable. 7 control(s) had complete mapped-assessor execution evidence; 3 lacked sufficient execution evidence and 0 were outside the current assessor mapping. Unmet controls have findings at or above the framework's mandatory severity threshold and should be remediated before relying on this server in a regulated deployment. All claims are traceable to individual finding rows via finding_id and to the governing rule via rule_id; the enclosing signed envelope commits MCP Sentinel to the exact bytes of this report.
Coverage & transparency
- Coverage band
- high
- Coverage ratio
- 88%
- Rules version
- 2026-04-23
Analysis techniques applied:
- ast-taint
- capability-graph
- composite
- cross-module
- dependency-audit
- linguistic
- schema-inference
- similarity
- structural
- stub
- unicode
Controls (10)
| ID | Control | Status | Evidence |
|---|---|---|---|
| MCP01 | Prompt Injection | ? Not assessed | 0 |
| MCP02 | Tool Poisoning | ✗ Unmet | 1 |
| MCP03 | Command Injection | ✓ Met | 0 |
| MCP04 | Data Exfiltration | ? Not assessed | 0 |
| MCP05 | Privilege Escalation | ✗ Unmet | 2 |
| MCP06 | Excessive Permissions | ✗ Unmet | 1 |
| MCP07 | Insecure Configuration | ✗ Unmet | 4 |
| MCP08 | Dependency Vulnerabilities | ✗ Unmet | 15 |
| MCP09 | Logging & Monitoring Failures | ✓ Met | 0 |
| MCP10 | Supply Chain Compromise | ? Not assessed | 0 |
Control details
MCP01 — Prompt Injection
? Not assessed22/24 mapped assessor rule(s) are proven to have executed; no adverse findings were observed, but the control is not assessed because execution evidence is incomplete. Unexecuted assessors: I3 (resource_templates, resources); I6 (prompts).
Control text:
Attacker-controlled content delivered through MCP server descriptions, parameters, resources, prompts, or initialize metadata that hijacks agent reasoning. Includes direct and indirect injection.
MCP02 — Tool Poisoning
✗ Unmet12/15 mapped assessor rule(s) are proven to have executed; 1 finding(s) observed (1 high); at least one finding is at or above the high threshold (status: unmet). 3 mapped assessor(s) did not execute, so this finding-backed result does not imply complete control coverage.
Control text:
Tools that misrepresent their behaviour through deceptive names, annotations, namespace squatting, homoglyphs, or time-based drift (rug pulls).
- High Excessive Scope Claims in Description (finding 5f148b3d-59d4-4ad5-983f-fcdbbbf2ea86, confidence 80%) SOURCE: external-content at tool execute — Tool description for "execute" advertises excessive scope via 1 claim pattern(s). Strongest: "all-data-scope". AI clients use the description to shape permis
- Rewrite the description so it accurately reflects the tool's scope. Replace sweeping claims ("full access", "unrestricted", "admin mode") with the specific operations the tool actually performs and the specific subset of resources it touches. The AI client uses the description as its mental model of acceptable privilege — overclaims directly undermine least-privilege review.
MCP03 — Command Injection
✓ Met6/6 mapped assessor rule(s) are proven to have executed; no adverse findings were observed on the mapped detector surface.
Control text:
Server code that executes shell commands, system calls, or git operations built from attacker-controlled input without sanitisation.
MCP04 — Data Exfiltration
? Not assessed14/15 mapped assessor rule(s) are proven to have executed; no adverse findings were observed, but the control is not assessed because execution evidence is incomplete. Unexecuted assessors: I13 (multi_server_tools).
Control text:
Server features that permit sensitive data to leave the trust boundary via HTTP, DNS, headers, URIs, or covert channels.
MCP05 — Privilege Escalation
✗ Unmet6/7 mapped assessor rule(s) are proven to have executed; 2 finding(s) observed (2 high); at least one finding is at or above the high threshold (status: unmet). 1 mapped assessor(s) did not execute, so this finding-backed result does not imply complete control coverage.
Control text:
Vertical privilege increases via JWT confusion, deserialization, evaluation of untrusted input, OAuth scope widening, or capability escalation post-initialization.
- High No Authentication on Network-Exposed Server (finding 02a80d24-a021-4949-b00e-83fe2f832e78, confidence 85%) SOURCE: file-content at http.js:176:20 — Bare `listen(port)` call with no host argument — defaults to 0.0.0.0 on most Node stacks, and no middleware registration naming an authentication construct was
- High No Authentication on Network-Exposed Server (finding 2663bb12-269b-474a-90f9-bcfbb2bf010e, confidence 85%) SOURCE: file-content at src/http.ts:215:18 — Bare `listen(port)` call with no host argument — defaults to 0.0.0.0 on most Node stacks, and no middleware registration naming an authentication construct
- Register an authentication middleware before exposing the server on a network interface. For express: `app.use(authMiddleware)` BEFORE any route registration; for fastify: `fastify.addHook('preHandler', authHook)`; for FastAPI: `Depends(get_current_user)` on every router. Acceptable auth strategies: JWT bearer (validated against an issuer + audience), OAuth 2.1, API key (rotated, expirable, validated with constant-time comparison), or mutual TLS. NEVER rely on a query-string token alone — query strings leak via reverse-proxy logs and browser history. If the MCP server is meant to be workstation-local, bind to 127.0.0.1 / localhost instead of 0.0.0.0 — most listen() / uvicorn.run() calls default to 0.0.0.0 if the host argument is omitted, so set the host EXPLICITLY.
MCP06 — Excessive Permissions
✗ Unmet4/7 mapped assessor rule(s) are proven to have executed; 1 finding(s) observed (1 high); at least one finding is at or above the medium threshold (status: unmet). 3 mapped assessor(s) did not execute, so this finding-backed result does not imply complete control coverage.
Control text:
Tools or roots claiming broader filesystem, network, or credential access than required for their stated function.
- High Excessive Scope Claims in Description (finding 5f148b3d-59d4-4ad5-983f-fcdbbbf2ea86, confidence 80%) SOURCE: external-content at tool execute — Tool description for "execute" advertises excessive scope via 1 claim pattern(s). Strongest: "all-data-scope". AI clients use the description to shape permis
- Rewrite the description so it accurately reflects the tool's scope. Replace sweeping claims ("full access", "unrestricted", "admin mode") with the specific operations the tool actually performs and the specific subset of resources it touches. The AI client uses the description as its mental model of acceptable privilege — overclaims directly undermine least-privilege review.
MCP07 — Insecure Configuration
✗ Unmet11/11 mapped assessor rule(s) are proven to have executed; 4 finding(s) observed (2 low, 2 high); at least one finding is at or above the high threshold (status: unmet).
Control text:
Servers shipped with insecure defaults: no authentication, wildcard CORS, HTTP transport, weak session tokens, disabled TLS validation, empty sandboxes.
- Low Schema Allows Unconstrained Additional Properties (finding 9b9c2db8-f1e1-4274-824a-650c47f47743, confidence 75%) SOURCE: user-parameter at tool execute — Tool "execute" accepts arbitrary extra keys at 1 point(s) in its input contract. The declared properties are validated; the handler may still read undeclared k
- Low Schema Allows Unconstrained Additional Properties (finding d2169aef-c097-4622-adfa-f9b996b1094e, confidence 75%) SOURCE: user-parameter at tool search_docs — Tool "search_docs" accepts arbitrary extra keys at 1 point(s) in its input contract. The declared properties are validated; the handler may still read unde
- High No Authentication on Network-Exposed Server (finding 02a80d24-a021-4949-b00e-83fe2f832e78, confidence 85%) SOURCE: file-content at http.js:176:20 — Bare `listen(port)` call with no host argument — defaults to 0.0.0.0 on most Node stacks, and no middleware registration naming an authentication construct was
- High No Authentication on Network-Exposed Server (finding 2663bb12-269b-474a-90f9-bcfbb2bf010e, confidence 85%) SOURCE: file-content at src/http.ts:215:18 — Bare `listen(port)` call with no host argument — defaults to 0.0.0.0 on most Node stacks, and no middleware registration naming an authentication construct
- Set additionalProperties: false on every object schema. This rejects any key outside the declared properties, closing the side-channel smuggling path and enforcing the schema's stated contract.
- Register an authentication middleware before exposing the server on a network interface. For express: `app.use(authMiddleware)` BEFORE any route registration; for fastify: `fastify.addHook('preHandler', authHook)`; for FastAPI: `Depends(get_current_user)` on every router. Acceptable auth strategies: JWT bearer (validated against an issuer + audience), OAuth 2.1, API key (rotated, expirable, validated with constant-time comparison), or mutual TLS. NEVER rely on a query-string token alone — query strings leak via reverse-proxy logs and browser history. If the MCP server is meant to be workstation-local, bind to 127.0.0.1 / localhost instead of 0.0.0.0 — most listen() / uvicorn.run() calls default to 0.0.0.0 if the host argument is omitted, so set the host EXPLICITLY.
MCP08 — Dependency Vulnerabilities
✗ Unmet7/7 mapped assessor rule(s) are proven to have executed; 15 finding(s) observed (3 high, 12 medium); at least one finding is at or above the high threshold (status: unmet).
Control text:
Known-CVE, abandoned, malicious, or typosquatted dependencies shipped with the server.
- High Known CVEs in Dependencies (finding 6a8f2b39-b87d-4150-9a7f-5eebdfebd915, confidence 92%) SOURCE: external-content at npm:qs@6.14.1 — Third-party package dependencies are external content resolved from public registries. A version with a published CVE ships the vulnerable code path as-is;
- High Known CVEs in Dependencies (finding 4ff249a2-d4c8-4242-b717-c65a98671943, confidence 78%) SOURCE: external-content at npm:@hono/node-server@1.19.10 — Third-party package dependencies are external content resolved from public registries. A version with a published CVE ships the vulnerable c
- High Known CVEs in Dependencies (finding 40d1a446-3075-4c90-b4c9-3245a41a892b, confidence 92%) SOURCE: external-content at npm:hono@4.12.4 — Third-party package dependencies are external content resolved from public registries. A version with a published CVE ships the vulnerable code path as-is
- Medium Abandoned Dependencies (finding f814e72b-9a41-4b83-bd73-a18e75df1bd6, confidence 70%) SOURCE: external-content at npm:zod@3.25.20 — An unmaintained dependency receives no patches for newly disclosed vulnerabilities. Every published CVE in the package after its last release date remains
- Medium Abandoned Dependencies (finding fd457992-35da-49d4-89bb-280b2fb9be7c, confidence 70%) SOURCE: external-content at npm:zod-to-json-schema@3.24.5 — An unmaintained dependency receives no patches for newly disclosed vulnerabilities. Every published CVE in the package after its last releas
- Medium Abandoned Dependencies (finding 4f6701ae-b83e-4bf6-8468-2fcc1a816882, confidence 70%) SOURCE: external-content at npm:zod-validation-error@4.0.1 — An unmaintained dependency receives no patches for newly disclosed vulnerabilities. Every published CVE in the package after its last relea
- Medium Abandoned Dependencies (finding 9189b613-655e-4cd6-a9ea-c0ab8433d49f, confidence 70%) SOURCE: external-content at npm:@cloudflare/cabidela@0.2.4 — An unmaintained dependency receives no patches for newly disclosed vulnerabilities. Every published CVE in the package after its last relea
- Medium Abandoned Dependencies (finding 90cf551b-ef44-4c18-8beb-c5208bbacf3f, confidence 70%) SOURCE: external-content at npm:@valtown/deno-http-worker@0.0.21 — An unmaintained dependency receives no patches for newly disclosed vulnerabilities. Every published CVE in the package after its last
- Medium Abandoned Dependencies (finding 2aa38ffd-c1e3-487c-8cb0-795262049bce, confidence 70%) SOURCE: external-content at npm:cookie-parser@1.4.6 — An unmaintained dependency receives no patches for newly disclosed vulnerabilities. Every published CVE in the package after its last release date
- Medium Abandoned Dependencies (finding b8bdb990-b13f-464b-99b3-d20e5577dec0, confidence 70%) SOURCE: external-content at npm:cors@2.8.5 — An unmaintained dependency receives no patches for newly disclosed vulnerabilities. Every published CVE in the package after its last release date remains
- Medium Abandoned Dependencies (finding 1f5b31b6-5f6a-47f0-9136-12d563a7e4df, confidence 70%) SOURCE: external-content at npm:express@5.1.0 — An unmaintained dependency receives no patches for newly disclosed vulnerabilities. Every published CVE in the package after its last release date remai
- Medium Abandoned Dependencies (finding bec8b394-2c91-4f0a-90a6-cc771396368e, confidence 70%) SOURCE: external-content at npm:fuse.js@7.1.0 — An unmaintained dependency receives no patches for newly disclosed vulnerabilities. Every published CVE in the package after its last release date remai
- Medium Abandoned Dependencies (finding 34dde0ca-675f-4e97-923e-2f508dfab390, confidence 70%) SOURCE: external-content at npm:minisearch@7.2.0 — An unmaintained dependency receives no patches for newly disclosed vulnerabilities. Every published CVE in the package after its last release date re
- Medium Abandoned Dependencies (finding 2913030b-a6c5-415e-9f76-181174c1f867, confidence 70%) SOURCE: external-content at npm:typescript@5.8.3 — An unmaintained dependency receives no patches for newly disclosed vulnerabilities. Every published CVE in the package after its last release date re
- Medium Abandoned Dependencies (finding 219c4319-7fca-4b8f-94b9-2f8481152228, confidence 70%) SOURCE: external-content at npm:yargs@17.7.2 — An unmaintained dependency receives no patches for newly disclosed vulnerabilities. Every published CVE in the package after its last release date remain
- Update the flagged dependency to a version that resolves every listed CVE. Consult the NVD or OSV advisory linked in the finding for the minimum safe version. Re-run `npm audit` / `pip-audit` / `osv-scanner` after the bump to confirm no residual advisories remain. Where a patched version is not yet released, apply an override (npm `overrides`, `resolutions`, `pnpm.overrides`, or a Python constraints file) to pin a compatible fix, and file a tracking issue so the override is removed when the upstream fix lands.
- Replace the abandoned dependency with an actively maintained alternative. Where no alternative exists, vendor the minimal code the project uses and delete the dependency. If the package is internal/private and legitimately stable, mark it as such in a repository-local allowlist so the scanner can skip it on future runs. Track the abandonment risk in the project's SBOM per ISO 27001 A.8.8.
MCP09 — Logging & Monitoring Failures
✓ Met6/6 mapped assessor rule(s) are proven to have executed; no adverse findings were observed on the mapped detector surface.
Control text:
Absent, insufficient, or tamperable audit logging that prevents incident detection and response.
MCP10 — Supply Chain Compromise
? Not assessed24/27 mapped assessor rule(s) are proven to have executed; no adverse findings were observed, but the control is not assessed because execution evidence is incomplete. Unexecuted assessors: I5 (resource_templates, resources); L1 (source_files(ci-config)); L3 (source_files(container-config)).
Control text:
Compromise of the build, publish, or distribution pipeline upstream of the server — including CI/CD token theft, build-artifact tampering, and manifest confusion.
Multi-step attack chains
No multi-step attack chains were synthesized for this server.
Cryptographic attestation
- Algorithm
- HMAC-SHA256
- Key ID
- mcps-hkdf-e8990a4e1526
- Signer
- mcp-sentinel/v1
- Signed at
- 2026-09-21T08:04:02.860Z
- Canonicalization
- RFC8785
HMAC-SHA256 signature (base64, wrapped at 64 chars):
ZvvQp7cDYzv8J+Cb2zv0+OKZEYJ/lHg5vIBVW7r0XRU=
Verification instructions:
To verify this report: 1. Extract the report body (everything except the .attestation field). 2. Canonicalize the body via RFC 8785 (JCS). 3. Compute HMAC-SHA256 with the signing key for key_id "mcps-hkdf-e8990a4e1526". 4. Base64-encode the result and compare with the signature above.