openai-anthropic-vendor-risk-v2
6 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.vendor.openai-realtime-without-auth-or-rate-limit
Section titled “seceng.vendor.openai-realtime-without-auth-or-rate-limit”info typescript, javascript, python openai, anthropic
OpenAI Realtime/WebRTC/WebSocket integration appears without visible auth, rate limit, session scoping, or ephemeral token controls.
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-306 CWE-770 LLM02 LLM06 LLM10 ASI03 ASI08 ASI02 MAP MANAGE
seceng.vendor.openai-assistants-thread-not-user-bound
Section titled “seceng.vendor.openai-assistants-thread-not-user-bound”info typescript, javascript, python openai, anthropic
OpenAI Assistants/Threads usage appears without visible user, tenant, or session binding.
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 LLM02 LLM07 LLM08 ASI03 ASI06 MAP MANAGE
seceng.vendor.openai-file-search-vector-store-no-boundary
Section titled “seceng.vendor.openai-file-search-vector-store-no-boundary”info typescript, javascript, python openai, anthropic
OpenAI file_search/vector store usage appears without visible tenant/user metadata, expiration, or file ownership controls.
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-639 CWE-200 LLM02 LLM07 LLM08 ASI03 ASI06 MAP MANAGE
seceng.vendor.openai-prompt-cache-sensitive-context
Section titled “seceng.vendor.openai-prompt-cache-sensitive-context”info typescript, javascript, python openai, anthropic
OpenAI/Anthropic prompt caching appears to include sensitive prompt, memory, transcript, or tenant context without visible cache policy.
Confidence P1: dangerous in context. False-positive risk high: it finds a lead to confirm, so expect benign hits.
Edge inspected: writes one thing writes to another.
CWE-200 LLM02 LLM07 LLM08 ASI03 ASI06 GOVERN MANAGE
seceng.vendor.anthropic-computer-use-without-domain-or-action-policy
Section titled “seceng.vendor.anthropic-computer-use-without-domain-or-action-policy”info typescript, javascript, python openai, anthropic
Anthropic computer_use/browser/computer tool appears enabled without visible domain, action, approval, or sandbox policy.
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-862 LLM06 ASI02 ASI10 GOVERN MANAGE
seceng.vendor-openai-tool-choice-required-dangerous-tools
Section titled “seceng.vendor-openai-tool-choice-required-dangerous-tools”info typescript, javascript, python openai, anthropic
OpenAI tool_choice appears forced/required while dangerous tools are available, without visible approval policy.
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).