ai-forensics-observability-v2
7 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.forensics.ai-action-missing-correlation-id
Section titled “seceng.forensics.ai-action-missing-correlation-id”info typescript, javascript, python, java, csharp, go, rust github-actions
AI action, tool call, or agent workflow lacks visible correlation/request/trace ID capture.
Confidence P1: dangerous in context. False-positive risk high: it finds a lead to confirm, so expect benign hits.
Edge inspected: none a property of one thing, not a relationship between two.
CWE-778 LLM07 LLM09 ASI02 ASI09 GOVERN MEASURE MANAGE
seceng.forensics.tool-call-missing-audit-event
Section titled “seceng.forensics.tool-call-missing-audit-event”info typescript, javascript, python, java, csharp, go, rust ai-forensics-observability
Tool/function call execution lacks visible security audit event with user, tenant, tool, args hash, result, and approval context.
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.
CWE-778 LLM06 LLM07 LLM09 ASI03 ASI02 ASI09 GOVERN MEASURE MANAGE
seceng.forensics-prompt-hash-missing
Section titled “seceng.forensics-prompt-hash-missing”info typescript, javascript, python, java, csharp, go, rust ai-forensics-observability
AI request appears to log or process prompt/messages without prompt hash or immutable evidence reference.
Confidence P1: dangerous in context. False-positive risk high: it finds a lead to confirm, so expect benign hits.
Edge inspected: none a property of one thing, not a relationship between two.
CWE-778 LLM07 LLM09 ASI01 ASI09 GOVERN MEASURE
seceng.forensics-model-version-missing
Section titled “seceng.forensics-model-version-missing”info typescript, javascript, python, java, csharp, go, rust ai-forensics-observability
AI call or model gateway invocation lacks visible model/version/provider capture for forensic reproducibility.
Confidence P1: dangerous in context. False-positive risk high: it finds a lead to confirm, so expect benign hits.
Edge inspected: none a property of one thing, not a relationship between two.
CWE-778 LLM09 LLM10 ASI09 ASI08 MAP MEASURE
seceng.forensics-ai-error-swallowed
Section titled “seceng.forensics-ai-error-swallowed”info typescript, javascript, python, java, csharp, go, rust ai-forensics-observability
AI/tool execution error appears swallowed or returned without structured security event capture.
Confidence P1: dangerous in context. False-positive risk high: it finds a lead to confirm, so expect benign hits.
Edge inspected: none a property of one thing, not a relationship between two.
CWE-778 CWE-391 LLM07 LLM10 ASI02 ASI08 MEASURE MANAGE
seceng.forensics-sensitive-ai-audit-raw
Section titled “seceng.forensics-sensitive-ai-audit-raw”warning typescript, javascript, python, java, csharp, go, rust ai-forensics-observability
AI forensic/audit log appears to store raw prompt, response, messages, or tool args instead of hashes/redacted evidence.
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: sends data leaves for another party; it crosses the secret->outside boundary (a secret leaving the trusted side).
CWE-532 CWE-200 LLM02 LLM07 ASI02 GOVERN MANAGE
seceng.forensics-no-user-tenant-on-ai-event
Section titled “seceng.forensics-no-user-tenant-on-ai-event”info typescript, javascript, python, java, csharp, go, rust ai-forensics-observability
AI security event or trace lacks visible user, tenant, workspace, organization, account, or session context.
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.
CWE-778 CWE-639 LLM07 LLM08 LLM09 ASI03 ASI06 ASI09 GOVERN MEASURE