agent-trust-assignment-v1
2 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.
keryn.trust.py.identity-assigned-from-model-output
Section titled “keryn.trust.py.identity-assigned-from-model-output”error python data-flow rule python, openai, anthropic, google-genai, langchain
A session, token, identity, privilege or trust field is assigned a value derived from model output or a tool/agent result. Model and tool output is attacker-influenced, so this lets a prompt or a poisoned tool result set the caller’s identity or privileges (confused deputy). Derive identity only from a verified principal; never from model or tool output.
Confidence P1: dangerous in context. 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-807 CWE-639 LLM06 ASI07 ASI03
Fix: Set identity, session, token, privilege and trust fields only from an authenticated principal (verified session, signed token, server-side lookup). Treat model and tool output as untrusted data — never assign it to a security-relevant field.
References: https://cwe.mitre.org/data/definitions/807.html · https://genai.owasp.org/resource/owasp-top-10-for-agentic-applications-for-2026/
Try it: a snippet that triggers this rule
Written for this page, not taken from any project. Save it as web/session_agent.py and scan it with keryn scan web --verbose; the finding below appears.
from anthropic import Anthropic
client = Anthropic()
def resolve_user(request, message_text: str): reply = client.messages.create( model="claude-sonnet-4-5", max_tokens=50, messages=[ {"role": "user", "content": f"Extract the account id mentioned here: {message_text}"} ], ) request.session.user_id = reply.content[0].text.strip() return request.session.user_idChecked with Keryn’s browser engine when this page was built: it reports keryn.trust.py.identity-assigned-from-model-output at line 14 (error, proof 2). The rule’s own logic is not published, so it cannot run in this tab; the pack list in the playground shows what does.
keryn.trust.ts.identity-assigned-from-model-output
Section titled “keryn.trust.ts.identity-assigned-from-model-output”error typescript, javascript data-flow rule typescript, nodejs, openai, anthropic, vercel-ai-sdk, langchain
A session, token, identity, privilege or trust field is assigned a value derived from model output or a tool/agent result. Model and tool output is attacker-influenced, so this lets a prompt or a poisoned tool result set the caller’s identity or privileges (confused deputy). Derive identity only from a verified principal.
Confidence P1: dangerous in context. 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-807 CWE-639 LLM06 ASI07 ASI03
Fix: Set identity, session, token, privilege and trust fields only from an authenticated principal; treat model and tool output as untrusted data.
References: https://cwe.mitre.org/data/definitions/807.html · https://genai.owasp.org/resource/owasp-top-10-for-agentic-applications-for-2026/