SECURITY REVIEW

Threat model and control evidence for MCP governance.

This page is the HTML entry for security reviewers. It summarizes the product threat model, the S1 to S31 promises, and dated control-suite records with their attestation status. Full tables and abuse-case detail stay in the linked rendered documents so evidence stays auditable.

BOUNDARY

What is under review.

Aegis is reviewed as the MCP decision plane after host identity is established and before a source presents a backend credential. Own-host and Knox are front-door choices; both feed the same governance engine.

Aegis threat model: trust zones Z1 to Z6, STRIDE highlights, and the governed MCP boundary before backend credentials.

In scope

Tool, resource, prompt, and completion calls; catalog pinning; approvals and receipts; egress; audit; hooks; output bounds; declarative HTTP and typed sources as capability only.

Out of scope

Model training and RAG poisoning, agents that never call Aegis, backend-internal ACLs beyond the gateway credential, discovery of shadow MCP servers off this path, and host OS or SIEM product choice.

Cross-zone rule: a call may not present a backend credential until the governance core has written ALLOW_INTENT (or denied). Caller tokens never enter the backend zone for protected backends.

TRUST ZONES

Six zones, one fail-closed decision plane.

ZoneComponentsPosture
Z1 ClientAgent, IDE, MCP client, untrusted documentsUntrusted for authorization
Z2 Host perimeterTLS, JWT/JWKS/static/SPIFFE, Knox Subject, CORS, routes, scopesTrusted only after verification
Z3 Governance coreCatalog, policy, egress, screening, approval, receipt, hooks, auditFail-closed MCP action root
Z4 SourcesHTTP, MCP proxy, pluginsCapability only; must not bypass Z3
Z5 BackendEngines that receive the gateway credentialTrusted for own responses; ACLs still required
Z6 Operator / IdP / networkConfig, pins, secrets, firewall, inventoryExternal residual owners
Aegis MCP-first governance core with own-host and optional Knox as deployment choices.

STRIDE AND ABUSE

Named abuse scenarios with expected denials.

The threat model elaborates STRIDE per zone into abuse cases TM01 to TM10. Each case states preconditions, attacker goal, expected gateway outcome, and the S-property that owns the proof.

TM01 Prompt-injected destructive call

Untrusted context steers a privileged caller. Expect policy, approval, and receipt binding before any backend write.

TM02 Tool poisoning / rug-pull

A changed tool definition is quarantined; prior approvals and receipts do not bind the new digest. With a signed serverCatalog, only the reviewed pins can bind and run-time acceptance is refused.

TM03 Confused deputy

Inbound caller tokens are not forwarded to protected backends; token exchange is egress-checked and subject-bound.

TM04 SSRF via nested URL

Nested HTTP(S) arguments and DNS answers pass egress classification; redirects are not followed.

TM05 READ carries a mutate

Classification, sql-readonly hooks, and least-privilege credentials contain the call; fail closed on ambiguity.

TM06 Self-approval

Requester cannot approve under an alternate spelling; delegated identities cannot approve.

TM07 Shadow MCP bypass

External residual: Aegis cannot govern a path that never reaches it. Network inventory remains operator work. For servers behind Aegis, a signed serverCatalog limits which may start, and a break-glass order can contain one during an incident.

TM08 to TM10

Post-delivery disclosure, supply-chain pin trust, and host CSRF/Origin checks are covered in the full model.

Read every TM01 to TM10 scenario with preconditions and expected outcomes.

OWASP MCP TOP 10

Coverage is not a claim of elimination.

Aegis coverage of all ten OWASP MCP risk areas, grouped as strong, partial, limited, or external.
MCP01 to MCP03Strong gateway controls for tokens, privilege scope, and changed tool definitions when the first pin is trusted.
MCP04 and MCP05Partial supply-chain and command-execution coverage; release controls and runtime sandboxing remain outside Aegis.
MCP06Prompt-injection screening is heuristic. Typed scope and independent approval are the durable safeguards.
MCP07 and MCP08Strong authentication/authorization and local audit evidence; IdP operations and SIEM remain deployment work.
MCP09Shadow servers cannot be discovered by a gateway that is not on their path. Behind Aegis, a signed server catalog limits which configured servers may start.
MCP10Caller-filtered discovery helps, but clients can retain or disclose received data.

PROMISES

Security properties S1 to S31.

Each property names the enforcing class and the test that must stay green. Breaking one is treated as a vulnerability. Highlights for reviewers:

S1 to S4Identity required, JWT rules, no token passthrough, default deny.
S5 to S7Approval, dry-run receipt, hash-chained audit intent.
S8 to S11Egress, catalog pin, request agreement, bounded redacted output.
Open the full S1 to S31 table

RECORDED EVIDENCE

Focused control-suite record.

The 2026-10-05 record reports a focused S1 to S31 selection run on JDK 17 with Maven Surefire: 227 selected tests, 0 failures. Tests added after that date are outside the 227 count.

A later hardening pass on 2026-11-07 ran the eight core modules as one build with every gate on: 408 tests, 0 failures, with SpotBugs and the FindSecBugs plugin, JaCoCo coverage, and the Apache RAT license check all green. The SpotBugs run reported 72 findings on its first pass; the real defects (constant-time digest comparison, a regex denial-of-service path, null guards, dead code) were fixed, and every remaining exclusion carries a written reason. CodeQL and an OSV dependency scan are committed as CI workflows but are not yet run from this host.

Build gates in the hardened profile: SpotBugs with FindSecBugs (static and security analysis), JaCoCo (coverage), Apache RAT (license headers), and the CI-side CodeQL and OSV scans. The CVE containment demo adds a runtime before-and-after proof against three real MCP server flaw classes.

Included

Governance, SoD, egress, catalog, admin CSRF, JWT/scopes, Knox packaging boundary, token exchange unit tests, output bounds, hooks fail-closed.

Separate evidence

Real-engine Keep runs, Keycloak token exchange, OpenBao secrets, and Postgres state have their own evidence files under docs/evidence/.

Read the security validation record and claims matrix G03.

RESIDUAL RISK

What operators still own.

Report vulnerabilities privately via a GitHub security advisory. Do not open a public issue for an unfixed vulnerability.

REFERENCE MAP

Continue in the full documents.