orm-rag-tenant-boundary-v2
9 rules · Keryn Software Licence · rules 0.8.0 · all packs
How to read a rule
- Severity
- How bad the finding is if it is real:
errorfails a default gate,warningandinfodo not. - Confidence
- P0 exploitable as written; P1 dangerous in context; P2 a heuristic or a posture check: confirm by reading the code.
- False-positive risk
- low: an exact API or flag, so hits are almost always real; medium: the pattern is legitimate in many programs, so check each hit; high: it finds a lead to confirm, so expect benign hits.
- Edge
- The relationship the finding sits on. reaches: content or traffic gets to a thing; calls: one thing invokes another; reads: one thing reads from another; writes: one thing writes to another; sends: data leaves for another party; grants: authority is given or left unchecked; depends_on: something relies on a component it does not verify; trusts: one party believes another without evidence. The boundary it crosses, when there is one: user->prompt (untrusted input reaching a prompt); model->code (model output reaching code, a shell, a query or a page); secret->outside (a secret leaving the trusted side); supply-chain (code, models or prompts pulled in from outside).
- Taxonomy chips
- References for readers who work in OWASP LLM 2025, OWASP Agentic, CWE or NIST AI RMF. They say where a rule belongs, not that a framework is satisfied.
seceng.orm.prisma-ai-resource-findunique-without-owner
Section titled “seceng.orm.prisma-ai-resource-findunique-without-owner”warning typescript, javascript orm
Prisma findUnique loads AI/RAG/chat/memory resource by ID without visible user, tenant, organization, or workspace ownership scope.
Confidence P2: a heuristic or a posture check: confirm by reading the code. False-positive risk medium: the pattern is legitimate in many programs, so check each hit.
Edge inspected: calls one thing invokes another; it crosses the model->code boundary (model output reaching code, a shell, a query or a page).
CWE-639 LLM06 LLM08 ASI03 ASI06 ASI02
seceng.orm.prisma-ai-findmany-without-tenant-filter
Section titled “seceng.orm.prisma-ai-findmany-without-tenant-filter”warning typescript, javascript orm
Prisma query reads AI/RAG/chat/memory records without visible user, tenant, organization, or workspace filter.
Confidence P1: dangerous in context. False-positive risk high: it finds a lead to confirm, so expect benign hits.
Edge inspected: grants authority is given or left unchecked.
seceng.orm.prisma-model-controlled-id-update-delete
Section titled “seceng.orm.prisma-model-controlled-id-update-delete”warning typescript, javascript orm
Prisma update/delete uses model or tool controlled ID without visible ownership check. Agentic tools can modify another tenant’s AI records.
Confidence P2: a heuristic or a posture check: confirm by reading the code. False-positive risk medium: the pattern is legitimate in many programs, so check each hit.
Edge inspected: grants authority is given or left unchecked.
CWE-639 LLM06 LLM08 ASI03 ASI02 ASI06
seceng.orm.prisma-ai-output-direct-write
Section titled “seceng.orm.prisma-ai-output-direct-write”warning typescript, javascript orm
AI-generated text or object is written directly to Prisma without visible schema validation or moderation.
Confidence P2: a heuristic or a posture check: confirm by reading the code. False-positive risk medium: the pattern is legitimate in many programs, so check each hit.
Edge inspected: writes one thing writes to another.
CWE-20 CWE-915 LLM08 LLM05 ASI06
seceng.orm.drizzle-ai-query-without-tenant-filter
Section titled “seceng.orm.drizzle-ai-query-without-tenant-filter”warning typescript, javascript orm
Drizzle query selects AI/RAG/chat/memory records without visible tenant/user/workspace predicate.
Confidence P1: dangerous in context. False-positive risk high: it finds a lead to confirm, so expect benign hits.
Edge inspected: grants authority is given or left unchecked.
seceng.orm.drizzle-model-controlled-id-write
Section titled “seceng.orm.drizzle-model-controlled-id-write”warning typescript, javascript orm
Drizzle update/delete uses model or tool controlled ID without visible ownership constraint.
Confidence P2: a heuristic or a posture check: confirm by reading the code. False-positive risk medium: the pattern is legitimate in many programs, so check each hit.
Edge inspected: grants authority is given or left unchecked.
CWE-639 LLM06 LLM08 ASI03 ASI02 ASI06
seceng.orm.typeorm-findone-ai-resource-without-owner
Section titled “seceng.orm.typeorm-findone-ai-resource-without-owner”warning typescript, javascript orm
TypeORM findOne/findOneBy loads AI/RAG/chat/memory resource by ID without visible user or tenant ownership condition.
Confidence P1: dangerous in context. False-positive risk high: it finds a lead to confirm, so expect benign hits.
Edge inspected: calls one thing invokes another; it crosses the model->code boundary (model output reaching code, a shell, a query or a page).
seceng.orm.sequelize-ai-resource-without-owner
Section titled “seceng.orm.sequelize-ai-resource-without-owner”warning typescript, javascript orm
Sequelize query loads AI/RAG/chat/memory resource by ID without visible tenant/user ownership condition.
Confidence P1: dangerous in context. False-positive risk high: it finds a lead to confirm, so expect benign hits.
Edge inspected: calls one thing invokes another; it crosses the model->code boundary (model output reaching code, a shell, a query or a page).
seceng.orm.sqlalchemy-ai-query-without-tenant-filter
Section titled “seceng.orm.sqlalchemy-ai-query-without-tenant-filter”warning python orm-rag-tenant-boundary
SQLAlchemy query over AI/RAG/chat/memory model lacks visible tenant/user ownership predicate.
Confidence P1: dangerous in context. False-positive risk high: it finds a lead to confirm, so expect benign hits.
Edge inspected: calls one thing invokes another; it crosses the model->code boundary (model output reaching code, a shell, a query or a page).