CWE is MITRE’s catalog of software weaknesses. 143 CWE entries appear in rule metadata (rules 0.8.0). The matrix lists all of them, most covered first; the 30 with the most rules also get a section with their rules. A row without a section links to the catalog filtered to that CWE.
This is an edge map, not a score. A count says how many rules look at a class of weakness and how sure they are, not how safe a system is; a class with few rules may simply need few, and no number here means a framework is satisfied. Confidence: P0 exploitable as written, P1 dangerous in context, P2 heuristic or posture. Edges are the relationships the rules inspect (reaches, calls, reads, writes, sends, grants, depends_on, trusts).
Coverage matrix
Section titled “Coverage matrix”| Entry | Rules | P0 | P1 | P2 | Edges |
|---|---|---|---|---|---|
| CWE-78 OS Command Injection | 136 | 32 | 73 | 31 | reaches 4 · calls 109 · sends 11 · grants 4 · depends_on 8 |
| CWE-200 Exposure of Sensitive Information to an Unauthorized Actor | 119 | 8 | 72 | 39 | reaches 19 · calls 10 · reads 12 · writes 5 · sends 38 · grants 34 · depends_on 1 |
| CWE-20 Improper Input Validation | 84 | 2 | 24 | 58 | reaches 30 · calls 22 · reads 9 · writes 12 · grants 3 · depends_on 2 · trusts 6 |
| CWE-94 Code Injection | 84 | 14 | 58 | 12 | reaches 4 · calls 70 · writes 2 · sends 1 · grants 2 · depends_on 4 · trusts 1 |
| CWE-829 Inclusion of Functionality from Untrusted Control Sphere | 80 | 10 | 44 | 26 | reaches 7 · calls 12 · writes 1 · sends 6 · grants 4 · depends_on 50 |
| CWE-22 Path Traversal | 78 | 0 | 59 | 19 | reaches 12 · calls 16 · reads 34 · writes 12 · grants 2 · depends_on 2 |
| CWE-79 Cross-site Scripting | 70 | 2 | 43 | 25 | calls 65 · writes 1 · sends 2 · grants 2 |
| CWE-639 Authorization Bypass Through User-Controlled Key | 63 | 0 | 24 | 39 | reaches 1 · calls 16 · writes 5 · grants 41 |
| CWE-862 Missing Authorization | 61 | 1 | 31 | 29 | reaches 1 · calls 9 · writes 1 · grants 50 |
| CWE-798 Use of Hard-coded Credentials | 56 | 3 | 21 | 32 | reaches 3 · calls 1 · reads 13 · sends 2 · grants 37 |
| CWE-918 Server-Side Request Forgery | 56 | 1 | 41 | 14 | reaches 41 · calls 9 · writes 1 · grants 5 |
| CWE-250 Execution with Unnecessary Privileges | 55 | 5 | 40 | 10 | calls 11 · writes 1 · grants 42 · depends_on 1 |
| CWE-306 Missing Authentication for Critical Function | 52 | 2 | 24 | 26 | reaches 14 · calls 3 · sends 1 · grants 30 · trusts 4 |
| CWE-502 Deserialization of Untrusted Data | 49 | 8 | 33 | 8 | reaches 1 · calls 3 · reads 1 · grants 1 · depends_on 43 |
| CWE-494 Download of Code Without Integrity Check | 47 | 5 | 30 | 12 | calls 7 · sends 1 · grants 3 · depends_on 36 |
| CWE-89 SQL Injection | 43 | 7 | 22 | 14 | calls 43 |
| CWE-532 Insertion of Sensitive Information into Log File | 40 | 1 | 20 | 19 | reaches 1 · calls 1 · reads 8 · writes 1 · sends 25 · grants 4 |
| CWE-1427 Improper Neutralization of Input Used for LLM Prompting | 37 | 0 | 25 | 12 | reaches 19 · calls 5 · writes 6 · grants 3 · trusts 4 |
| CWE-770 Allocation of Resources Without Limits or Throttling | 36 | 1 | 13 | 22 | reaches 3 · calls 6 · writes 3 · sends 1 · grants 23 |
| CWE-345 Insufficient Verification of Data Authenticity | 32 | 7 | 15 | 10 | reaches 2 · calls 4 · writes 5 · grants 5 · depends_on 3 · trusts 13 |
| CWE-312 Cleartext Storage of Sensitive Information | 29 | 4 | 12 | 13 | reads 1 · writes 8 · sends 9 · grants 11 |
| CWE-73 External Control of File Name or Path | 28 | 0 | 18 | 10 | reaches 2 · calls 9 · reads 8 · writes 9 |
| CWE-506 Embedded Malicious Code | 26 | 14 | 10 | 2 | calls 9 · writes 5 · sends 11 · depends_on 1 |
| CWE-346 Origin Validation Error | 24 | 2 | 12 | 10 | reaches 13 · grants 11 |
| CWE-522 Insufficiently Protected Credentials | 23 | 7 | 12 | 4 | sends 1 · grants 22 |
| CWE-77 Command Injection | 22 | 0 | 14 | 8 | reaches 15 · calls 2 · grants 5 |
| CWE-116 Improper Encoding or Escaping of Output | 21 | 2 | 14 | 5 | reaches 6 · calls 7 · sends 3 · grants 5 |
| CWE-915 | 20 | 0 | 4 | 16 | calls 6 · reads 2 · writes 9 · grants 3 |
| CWE-327 | 18 | 0 | 16 | 2 | sends 7 · depends_on 1 · trusts 9 · none 1 |
| CWE-668 Exposure of Resource to Wrong Sphere | 18 | 3 | 10 | 5 | reaches 7 · calls 1 · grants 10 |
| CWE-269 | 17 | 2 | 13 | 2 | calls 5 · writes 1 · grants 11 |
| CWE-201 | 16 | 1 | 12 | 3 | reads 1 · writes 1 · sends 14 |
| CWE-400 | 16 | 1 | 9 | 6 | calls 4 · writes 3 · grants 9 |
| CWE-601 | 16 | 0 | 15 | 1 | reaches 9 · calls 6 · writes 1 |
| CWE-95 | 15 | 5 | 10 | 0 | calls 13 · depends_on 2 |
| CWE-319 | 15 | 0 | 12 | 3 | sends 12 · grants 1 · depends_on 2 |
| CWE-1007 | 15 | 2 | 10 | 3 | calls 9 · grants 5 · depends_on 1 |
| CWE-88 | 14 | 9 | 4 | 1 | calls 14 |
| CWE-693 Protection Mechanism Failure | 14 | 0 | 12 | 2 | calls 2 · grants 10 · depends_on 1 · trusts 1 |
| CWE-1327 | 13 | 2 | 5 | 6 | reaches 11 · grants 2 |
| CWE-451 | 12 | 0 | 10 | 2 | calls 4 · grants 8 |
| CWE-778 | 12 | 0 | 7 | 5 | calls 2 · reads 1 · grants 5 · none 4 |
| CWE-295 | 11 | 0 | 11 | 0 | sends 1 · grants 3 · depends_on 2 · trusts 5 |
| CWE-441 | 11 | 0 | 11 | 0 | calls 4 · writes 1 · grants 6 |
| CWE-732 Incorrect Permission Assignment for Critical Resource | 11 | 1 | 9 | 1 | reaches 3 · calls 4 · grants 4 |
| CWE-1188 | 11 | 0 | 3 | 8 | grants 1 · trusts 10 |
| CWE-266 | 10 | 0 | 6 | 4 | calls 1 · reads 1 · grants 7 · depends_on 1 |
| CWE-347 | 10 | 9 | 1 | 0 | calls 1 · depends_on 1 · trusts 8 |
| CWE-287 | 8 | 2 | 3 | 3 | calls 3 · grants 4 · depends_on 1 |
| CWE-1240 | 8 | 0 | 8 | 0 | trusts 8 |
| CWE-1426 Improper Validation of Generative AI Output | 8 | 0 | 5 | 3 | calls 3 · sends 1 · trusts 4 |
| CWE-74 | 7 | 0 | 5 | 2 | reaches 6 · grants 1 |
| CWE-209 | 7 | 0 | 4 | 3 | reaches 2 · calls 2 · sends 3 |
| CWE-352 | 7 | 0 | 2 | 5 | calls 6 · grants 1 |
| CWE-470 | 7 | 1 | 5 | 1 | calls 5 · depends_on 2 |
| CWE-636 | 7 | 0 | 4 | 3 | calls 1 · grants 6 |
| CWE-912 | 7 | 5 | 2 | 0 | calls 1 · writes 6 |
| CWE-1336 | 7 | 3 | 4 | 0 | calls 7 |
| CWE-285 | 6 | 0 | 5 | 1 | calls 1 · grants 5 |
| CWE-754 | 6 | 0 | 4 | 2 | reads 4 · trusts 2 |
| CWE-1357 | 6 | 1 | 5 | 0 | calls 1 · depends_on 5 |
| CWE-15 | 5 | 0 | 5 | 0 | reaches 1 · calls 4 |
| CWE-280 | 5 | 0 | 4 | 1 | grants 5 |
| CWE-311 | 5 | 1 | 0 | 4 | reaches 3 · calls 1 · writes 1 |
| CWE-359 | 5 | 0 | 4 | 1 | calls 1 · sends 4 |
| CWE-427 | 5 | 1 | 0 | 4 | calls 4 · depends_on 1 |
| CWE-749 Exposed Dangerous Method or Function | 5 | 0 | 2 | 3 | reaches 4 · calls 1 |
| CWE-755 | 5 | 0 | 5 | 0 | calls 2 · grants 3 |
| CWE-943 | 5 | 0 | 4 | 1 | calls 5 |
| CWE-1021 | 5 | 0 | 2 | 3 | grants 2 · trusts 3 |
| CWE-1039 | 5 | 1 | 4 | 0 | trusts 5 |
| CWE-1333 | 5 | 0 | 3 | 2 | calls 5 |
| CWE-120 | 4 | 0 | 4 | 0 | grants 4 |
| CWE-179 | 4 | 0 | 4 | 0 | trusts 4 |
| CWE-180 | 4 | 0 | 4 | 0 | trusts 4 |
| CWE-611 | 4 | 0 | 3 | 1 | reaches 3 · calls 1 |
| CWE-807 Reliance on Untrusted Inputs in a Security Decision | 4 | 0 | 3 | 1 | calls 2 · grants 2 |
| CWE-1104 | 4 | 0 | 3 | 1 | depends_on 4 |
| CWE-1220 Insufficient Granularity of Access Control | 4 | 0 | 4 | 0 | grants 4 |
| CWE-259 | 3 | 2 | 0 | 1 | grants 3 |
| CWE-281 | 3 | 0 | 0 | 3 | grants 3 |
| CWE-290 | 3 | 0 | 2 | 1 | calls 3 |
| CWE-300 | 3 | 0 | 3 | 0 | sends 3 |
| CWE-320 | 3 | 0 | 0 | 3 | reaches 3 |
| CWE-362 | 3 | 0 | 0 | 3 | writes 3 |
| CWE-367 | 3 | 0 | 0 | 3 | writes 3 |
| CWE-390 Detection of Error Condition Without Action | 3 | 0 | 2 | 1 | calls 3 |
| CWE-454 | 3 | 0 | 3 | 0 | calls 3 |
| CWE-625 | 3 | 0 | 2 | 1 | calls 3 |
| CWE-776 | 3 | 0 | 1 | 2 | calls 2 · depends_on 1 |
| CWE-863 Incorrect Authorization | 3 | 0 | 2 | 1 | grants 3 |
| CWE-942 | 3 | 0 | 3 | 0 | grants 3 |
| CWE-203 | 2 | 0 | 0 | 2 | sends 2 |
| CWE-284 Improper Access Control | 2 | 0 | 2 | 0 | calls 1 · grants 1 |
| CWE-294 | 2 | 0 | 0 | 2 | trusts 2 |
| CWE-326 | 2 | 0 | 2 | 0 | calls 1 · sends 1 |
| CWE-328 | 2 | 0 | 1 | 1 | calls 1 · sends 1 |
| CWE-338 | 2 | 0 | 1 | 1 | calls 1 · sends 1 |
| CWE-377 | 2 | 0 | 1 | 1 | writes 1 · grants 1 |
| CWE-384 | 2 | 0 | 2 | 0 | reaches 2 |
| CWE-526 | 2 | 0 | 2 | 0 | grants 2 |
| CWE-598 | 2 | 0 | 2 | 0 | reaches 2 |
| CWE-653 Improper Isolation or Compartmentalization | 2 | 0 | 1 | 1 | grants 2 |
| CWE-757 | 2 | 0 | 2 | 0 | sends 2 |
| CWE-799 | 2 | 0 | 0 | 2 | trusts 2 |
| CWE-834 | 2 | 0 | 2 | 0 | grants 2 |
| CWE-840 | 2 | 2 | 0 | 0 | calls 2 |
| CWE-1059 | 2 | 0 | 0 | 2 | calls 2 |
| CWE-1321 | 2 | 0 | 1 | 1 | calls 1 · writes 1 |
| CWE-117 | 1 | 0 | 0 | 1 | reads 1 |
| CWE-125 | 1 | 0 | 1 | 0 | calls 1 |
| CWE-134 | 1 | 1 | 0 | 0 | calls 1 |
| CWE-155 | 1 | 0 | 1 | 0 | calls 1 |
| CWE-185 | 1 | 0 | 1 | 0 | grants 1 |
| CWE-208 | 1 | 0 | 0 | 1 | sends 1 |
| CWE-242 | 1 | 0 | 1 | 0 | grants 1 |
| CWE-322 | 1 | 0 | 1 | 0 | depends_on 1 |
| CWE-330 | 1 | 0 | 0 | 1 | calls 1 |
| CWE-379 | 1 | 0 | 0 | 1 | grants 1 |
| CWE-391 | 1 | 0 | 1 | 0 | none 1 |
| CWE-409 | 1 | 0 | 0 | 1 | writes 1 |
| CWE-459 | 1 | 1 | 0 | 0 | calls 1 |
| CWE-472 | 1 | 0 | 0 | 1 | calls 1 |
| CWE-489 | 1 | 0 | 1 | 0 | grants 1 |
| CWE-497 | 1 | 0 | 0 | 1 | reaches 1 |
| CWE-552 | 1 | 0 | 1 | 0 | calls 1 |
| CWE-602 | 1 | 0 | 0 | 1 | grants 1 |
| CWE-610 | 1 | 0 | 1 | 0 | reaches 1 |
| CWE-617 Reachable Assertion | 1 | 0 | 0 | 1 | calls 1 |
| CWE-665 | 1 | 0 | 1 | 0 | writes 1 |
| CWE-696 Incorrect Behavior Order | 1 | 0 | 0 | 1 | grants 1 |
| CWE-706 | 1 | 0 | 0 | 1 | depends_on 1 |
| CWE-780 | 1 | 0 | 1 | 0 | calls 1 |
| CWE-787 | 1 | 0 | 1 | 0 | calls 1 |
| CWE-835 | 1 | 0 | 0 | 1 | grants 1 |
| CWE-838 | 1 | 0 | 1 | 0 | reaches 1 |
| CWE-908 | 1 | 0 | 1 | 0 | writes 1 |
| CWE-916 | 1 | 0 | 1 | 0 | sends 1 |
| CWE-1004 | 1 | 0 | 1 | 0 | calls 1 |
| CWE-1088 | 1 | 0 | 1 | 0 | grants 1 |
| CWE-1287 | 1 | 0 | 1 | 0 | calls 1 |
| CWE-1295 | 1 | 0 | 1 | 0 | grants 1 |
| CWE-1395 | 1 | 0 | 1 | 0 | depends_on 1 |
CWE-78 OS Command Injection
Section titled “CWE-78 OS Command Injection”136 rules · confidence P0 32, P1 73, P2 31 · severity error 85, warning 42, info 9
Edges inspected: reaches 4 · calls 109 · sends 11 · grants 4 · depends_on 8
30 of 136 rules, surest first
- keryn.agenthooks.copilot-hook-pipes-download-to-shell
- keryn.agenthooks.copilot-hook-reverse-shell
- keryn.agenthooks.cursor-hook-pipes-download-to-shell
- keryn.agenthooks.cursor-hook-reverse-shell
- keryn.agenthooks.gemini-hook-pipes-download-to-shell
- keryn.agenthooks.gemini-hook-reverse-shell
- keryn.agenthooks.windsurf-hook-pipes-download-to-shell
- keryn.agenthooks.windsurf-hook-reverse-shell
- keryn.general.ruby.command-injection-request
- keryn.rustagent.model-output-to-process-command-rs
- keryn.skillexec.reverse-shell-bash-devtcp
- keryn.skillexec.reverse-shell-netcat
- keryn.skillexec.reverse-shell-python
- keryn.skillexec.reverse-shell-script-langs
- keryn.skillexec.reverse-shell-socat
- seceng.agentcfg.hook-pipes-to-shell
- seceng.llm05.taint.model-output-to-async-subprocess-python
- seceng.llm05.taint.model-output-to-child-process-js
- seceng.llm05.taint.model-output-to-code-interpreter-tool-python
- seceng.llm05.taint.model-output-to-mcp-stdio-command-js
- seceng.llm05.taint.model-output-to-os-command-python
- seceng.llm05.taint.model-output-to-remote-command-python
- seceng.llm05.taint.model-output-to-shell-quote-missing-js
- seceng.mcp.csharp.tool-process-start
- seceng.mcp.go.tool-exec-command
- seceng.mcp.go.tool-exec-shell-c
- seceng.mcp.java.tool-process-exec
- seceng.mcp.kotlin.tool-process-builder
- seceng.mcp.py.stdio-server-command-from-request
- seceng.mcp.rust.tool-command-new
CWE-200 Exposure of Sensitive Information to an Unauthorized Actor
Section titled “CWE-200 Exposure of Sensitive Information to an Unauthorized Actor”119 rules · confidence P0 8, P1 72, P2 39 · severity error 22, warning 81, info 16
Edges inspected: reaches 19 · calls 10 · reads 12 · writes 5 · sends 38 · grants 34 · depends_on 1
30 of 119 rules, surest first
- keryn.skillcred.browser-credential-store-read
- keryn.skillcred.macos-keychain-secret-read
- keryn.skillcred.os-credential-database-access
- keryn.skillcred.secret-manager-or-keyring-export
- keryn.skillcred.ssh-key-material-read
- seceng.agentrules.secrets-to-url-command
- seceng.llm-frameworks.dspy-signature-leakage
- seceng.llm02-system-prompt-in-client
- keryn.egress.cs.secret-to-side-channel
- keryn.egress.go.secret-to-side-channel
- keryn.egress.java.secret-to-side-channel
- keryn.egress.php.secret-to-side-channel
- keryn.egress.py.secret-to-side-channel
- keryn.egress.ts.secret-to-side-channel
- keryn.outbound.model-output-as-message-recipient-js
- keryn.outbound.model-output-as-message-recipient-py
- keryn.skillexec.clipboard-exfil
- keryn.skillpersist.dns-lookup-exfil-text
- seceng.ciagent.agent-step-env-secrets
- seceng.mcp.csharp.tool-file-read
- seceng.mcp.go.tool-file-read
- seceng.mcp.java.tool-files-read
- seceng.mcp.kotlin.tool-file-access
- seceng.mcp.rust.tool-fs-read
- keryn.agentcfg.env-api-base-url-override
- keryn.agentcfg.hook-command-dns-lookup-with-variable
- keryn.agenthooks.copilot-hook-sends-host-data
- keryn.agenthooks.copilot-prompt-hook-posts-to-network
- keryn.agenthooks.cursor-hook-sends-host-data
- keryn.agenthooks.cursor-prompt-hook-posts-to-network
CWE-20 Improper Input Validation
Section titled “CWE-20 Improper Input Validation”84 rules · confidence P0 2, P1 24, P2 58 · severity error 8, warning 67, info 9
Edges inspected: reaches 30 · calls 22 · reads 9 · writes 12 · grants 3 · depends_on 2 · trusts 6
30 of 84 rules, surest first
- keryn.outbound.model-output-as-payment-amount-js
- keryn.outbound.model-output-as-payment-amount-py
- keryn.canon.go.validate-before-decode
- keryn.canon.php.validate-before-decode
- keryn.canon.py.validate-before-decode
- keryn.canon.ts.validate-before-decode
- seceng.asi06.langgraph-langchain-memory-write-from-untrusted-python
- seceng.llm01.prompt-injection-taint
- keryn.outbound.model-output-to-crm-record-js
- keryn.outbound.model-output-to-crm-record-py
- seceng.github-actions-secrets-to-issue-comment-llm
- seceng.gitlab-ci-user-input-to-shell-or-ai
- seceng.llm-frameworks.js-ts-langchain-prompt-injection-validation-disabled
- seceng.llm-frameworks.python-langchain-prompt-injection-validation-disabled
- seceng.llm01.js-ts-concat-prompt
- seceng.llm01.prompt-injection-taint-js
- seceng.llm01.taint.tool-result-to-tool-args-js
- seceng.llm05.taint.model-output-to-memory-write-js
- seceng.llm06.langchain-tool-missing-args-schema
- seceng.mongoose-ai-regex-query-from-model-output
- seceng.schema-model-score-controls-security-decision
- seceng.supabase.rpc-query-text-directly-from-request
- seceng.tsorm-ai-generated-orderby-or-select
- seceng.github-ai-pr-bot-commits-without-approval
- seceng.langgraph.stategraph-without-typed-state
- seceng.llm01.trusted-dev-prompt-suppression
- keryn.agents2026.agentcore-entrypoint-passes-payload-unvalidated
- keryn.agents2026.genkit-definetool-no-inputschema
- keryn.agents2026.mastra-createtool-no-inputschema
- keryn.agents2026.vercelai-tool-permissive-input-schema
CWE-94 Code Injection
Section titled “CWE-94 Code Injection”84 rules · confidence P0 14, P1 58, P2 12 · severity error 61, warning 23, info 0
Edges inspected: reaches 4 · calls 70 · writes 2 · sends 1 · grants 2 · depends_on 4 · trusts 1
30 of 84 rules, surest first
- keryn.browseragent.model-output-to-page-script-js
- keryn.browseragent.model-output-to-page-script-py
- keryn.rustagent.model-output-to-webview-eval-rs
- seceng.ciagent.agent-pr-head-checkout-on-pull-request-target
- seceng.llm05.taint.model-output-to-browser-eval-js
- seceng.llm05.taint.model-output-to-code-interpreter-tool-python
- seceng.llm05.taint.model-output-to-dynamic-import-python
- seceng.llm05.taint.model-output-to-eval-js
- seceng.llm05.taint.model-output-to-eval-python
- seceng.llm05.taint.model-output-to-require-import-js
- seceng.llm05.taint.model-output-to-template-engine-js
- seceng.llm05.taint.model-output-to-template-python
- seceng.mcp.py.stdio-server-command-from-request
- seceng.mcp.ts.stdio-client-transport-command-from-request
- keryn.agents2026.agentcore-payload-to-code-interpreter
- keryn.agents2026.claude-agent-sdk-hook-runs-shell-from-tool-input
- keryn.agents2026.genkit-definetool-handler-command-exec
- keryn.agents2026.google-adk-unsafe-local-code-executor
- keryn.agents2026.openai-agents-function-tool-command-exec
- keryn.agents2026.openai-agents-js-tool-execute-command-exec
- keryn.agents2026.pydantic-ai-tool-argument-to-exec
- keryn.general.python.logging-config-listen
- keryn.general.ruby.render-request-to-template
- keryn.vfgap.script-engine-eval-java
- keryn.vfgap.vm-run-dynamic-code-js
- seceng.agentcfg.enable-all-project-mcp-servers
- seceng.agentrules.base64-decode-exec
- seceng.asi07.agent-event-stream-dispatches-commands-ts
- seceng.ciagent.agent-prompt-interpolates-event-text
- seceng.ciagent.agent-step-on-pull-request-target
CWE-829 Inclusion of Functionality from Untrusted Control Sphere
Section titled “CWE-829 Inclusion of Functionality from Untrusted Control Sphere”80 rules · confidence P0 10, P1 44, P2 26 · severity error 25, warning 50, info 5
Edges inspected: reaches 7 · calls 12 · writes 1 · sends 6 · grants 4 · depends_on 50
30 of 80 rules, surest first
- keryn.agenthooks.copilot-hook-pipes-download-to-shell
- keryn.agenthooks.cursor-hook-pipes-download-to-shell
- keryn.agenthooks.gemini-hook-pipes-download-to-shell
- keryn.agenthooks.windsurf-hook-pipes-download-to-shell
- keryn.modelload.torch-hub-load-trust-repo-nonliteral-py
- keryn.skillpersist.shell-rc-download-exec
- seceng.agentcfg.hook-pipes-to-shell
- seceng.ciagent.agent-pr-head-checkout-on-pull-request-target
- seceng.llm05.taint.model-output-to-mcp-stdio-command-js
- seceng.llm05.taint.model-output-to-require-import-js
- keryn.skillexec.download-chmod-run
- keryn.skillexec.download-to-temp-exec
- keryn.skillexec.encrypted-archive-exec
- keryn.skillexec.fetch-to-interpreter
- keryn.skillexec.staging-host-exec
- seceng.agentcfg.enable-all-project-mcp-servers
- seceng.agentcfg.mcp-npx-unpinned-package
- seceng.agentcfg.mcp-uvx-unpinned-package
- seceng.llm03.js-ts-npm-install-in-code
- seceng.mcp.config.server-package-lookalike-official-scope
- seceng.modelsc.hf-inference-client-user-model
- seceng.modelsc.ollama-pull-user-model
- seceng.modelsc.ollama-pull-user-model-ts
- seceng.modelsc.transformers-user-controlled-model-id
- seceng.modelsc.transformersjs-user-controlled-model-id
- keryn.agenthooks.copilot-hook-edits-workflows
- keryn.agenthooks.copilot-hook-raw-ip-url
- keryn.agenthooks.cursor-hook-raw-ip-url
- keryn.agenthooks.gemini-hook-raw-ip-url
- keryn.agenthooks.windsurf-hook-raw-ip-url
CWE-22 Path Traversal
Section titled “CWE-22 Path Traversal”78 rules · confidence P0 0, P1 59, P2 19 · severity error 41, warning 34, info 3
Edges inspected: reaches 12 · calls 16 · reads 34 · writes 12 · grants 2 · depends_on 2
30 of 78 rules, surest first
- keryn.general.csharp.path-traversal-request-to-file
- keryn.general.csharp.zip-slip-entry-name
- keryn.general.go.path-traversal-request-to-file
- keryn.general.go.zip-slip-entry-name
- keryn.general.java.path-traversal-request-to-file
- keryn.general.java.zip-slip-entry-name
- keryn.general.js.path-traversal-request-to-file
- keryn.general.kotlin.path-traversal-request-to-file
- keryn.general.kotlin.zip-slip-entry-name
- keryn.general.php.path-traversal-request-to-file
- keryn.general.php.zip-slip-archive-entry-name
- keryn.general.python.path-traversal-request-to-file
- keryn.general.ruby.file-access-request
- keryn.general.swift.path-traversal-request-to-file
- keryn.modelload.request-path-to-model-loader-python
- keryn.rustagent.model-output-to-fs-path-rs
- seceng.mcp.csharp.tool-file-read
- seceng.mcp.csharp.tool-file-write
- seceng.mcp.go.tool-file-delete
- seceng.mcp.go.tool-file-read
- seceng.mcp.go.tool-file-write
- seceng.mcp.java.tool-files-read
- seceng.mcp.java.tool-files-write
- seceng.mcp.kotlin.tool-file-access
- seceng.mcp.rust.tool-fs-read
- seceng.mcp.rust.tool-fs-remove
- seceng.mcp.rust.tool-fs-write
- seceng.modelsc.model-file-user-controlled-path
- seceng.ragloader.directory-loader-path-traversal
- seceng.ragloader.langchain-file-loader-path-traversal
CWE-79 Cross-site Scripting
Section titled “CWE-79 Cross-site Scripting”70 rules · confidence P0 2, P1 43, P2 25 · severity error 7, warning 61, info 2
Edges inspected: calls 65 · writes 1 · sends 2 · grants 2
30 of 70 rules, surest first
- keryn.rustagent.model-output-to-webview-eval-rs
- seceng.llm05.taint.model-output-to-browser-eval-js
- keryn.general.php.xss-request-to-output
- keryn.general.ruby.html-safe-request
- vulnfix-ai-ai_js_dangerous_html
- vulnfix-ai-ai_streamlit_unsafe_html
- vulnfix-ai-ai_taint_html
- keryn.browseragent.model-output-to-page-html-js
- keryn.browseragent.model-output-to-page-html-py
- keryn.general.csharp.xss-request-to-html-output
- keryn.general.go.xss-request-to-response-writer
- keryn.general.java.xss-request-to-response-writer
- keryn.general.js.template-escape-markup-disabled
- keryn.general.js.xss-request-to-html-response
- keryn.general.kotlin.xss-request-to-html-response
- keryn.general.python.jinja2-autoescape-disabled
- keryn.general.python.markup-non-literal
- keryn.general.python.xss-request-to-html-response
- keryn.general.ruby.render-inline-unescaped-local
- keryn.general.swift.webview-html-js-injection
- keryn.rails.cve.content-tag-attribute-value-cve-2016-6316
- keryn.rails.cve.mail-to-javascript-encode-cve-2011-0446
- keryn.rails.cve.number-helper-format-cve-2014-0081
- keryn.rails.cve.raw-sanitized-request-old-sanitizer
- keryn.rails.cve.sanitize-call-cve-2013-1857
- keryn.rails.cve.sanitize-in-view-cve-2013-1857
- keryn.rails.cve.sanitizer-select-style-cve-2022-32209
- keryn.rails.cve.select-options-string-cve-2012-1099
- keryn.rails.cve.select-tag-prompt-cve-2012-3463
- keryn.rails.cve.simple-format-cve-2013-6416
CWE-639 Authorization Bypass Through User-Controlled Key
Section titled “CWE-639 Authorization Bypass Through User-Controlled Key”63 rules · confidence P0 0, P1 24, P2 39 · severity error 7, warning 46, info 10
Edges inspected: reaches 1 · calls 16 · writes 5 · grants 41
30 of 63 rules, surest first
- keryn.trust.py.identity-assigned-from-model-output
- keryn.trust.ts.identity-assigned-from-model-output
- seceng.asi03.impersonation-header-from-model-output-python
- seceng.asi03.impersonation-header-from-model-output-ts
- seceng.asi03.tool-authz-from-model-argument-python
- seceng.asi06.memory-shared-user-scope-python
- seceng.asi06.memory-shared-user-scope-ts
- seceng.cloudflare-d1-ai-query-without-tenant-filter
- seceng.cloudflare-kv-ai-memory-without-tenant-prefix
- seceng.deno.edge-kv-ai-memory-without-tenant-prefix
- seceng.firebase.firestore-ai-collection-read-without-owner-filter
- seceng.mongoose.ai-query-without-tenant-filter
- seceng.orm.drizzle-ai-query-without-tenant-filter
- seceng.orm.prisma-ai-findmany-without-tenant-filter
- seceng.orm.sequelize-ai-resource-without-owner
- seceng.orm.sqlalchemy-ai-query-without-tenant-filter
- seceng.orm.typeorm-findone-ai-resource-without-owner
- seceng.supabase-auth-user-not-bound-to-query
- seceng.agentframework.autogen-shared-workdir-or-history
- seceng.agentframework.crewai-memory-without-tenant-scope
- seceng.forensics-no-user-tenant-on-ai-event
- seceng.redis-embedding-cache-shared-key
- seceng.tsorm-prisma-include-cross-tenant-ai-data
- seceng.vectordb.shared-collection-name
- keryn.privilege.client-field-trusted
- seceng.asi03.tool-authz-from-model-argument-ts
- seceng.cloudflare.vectorize-query-without-namespace-filter
- seceng.firebase.firestore-doc-id-from-request-without-owner
- seceng.graphql-resolver-id-from-args-without-owner
- seceng.java-ai-resource-by-id-without-owner
CWE-862 Missing Authorization
Section titled “CWE-862 Missing Authorization”61 rules · confidence P0 1, P1 31, P2 29 · severity error 9, warning 42, info 10
Edges inspected: reaches 1 · calls 9 · writes 1 · grants 50
30 of 61 rules, surest first
- seceng.workflow.webhook-to-credentialed-action
- keryn.agents2026.claude-agent-sdk-bypass-permissions
- keryn.agents2026.claude-agent-sdk-ts-bypass-permissions
- keryn.agents2026.strands-bypass-tool-consent
- seceng.agentcfg.default-mode-bypass-permissions
- seceng.asi09.approval-callback-always-allows-python
- seceng.asi09.approval-callback-always-allows-ts
- seceng.asi09.approval-gate-disabled-python
- seceng.asi09.approval-gate-disabled-ts
- seceng.ciagent.agent-skip-permissions-flag
- keryn.agentcfg.hook-command-unconditional-allow
- keryn.agentclient.codex-approval-policy-never
- keryn.agentclient.gemini-auto-accept
- keryn.agentclient.gemini-mcp-server-trusted
- keryn.agentclient.skill-allowed-tools-unrestricted
- keryn.agentclient.vscode-chat-tools-auto-approve
- keryn.agenthooks.cursor-gate-hook-always-allows
- keryn.agents2026.claude-agent-sdk-can-use-tool-always-allows
- keryn.agents2026.vercelai-tool-needsapproval-disabled-destructive
- keryn.gaps.money-moving-call-without-approval
- keryn.rails.ruby.skip-auth-filter-except
- keryn.vendor.openai-action-mutating-op-not-consequential-yaml
- seceng.agentrules.disable-permission-checks
- seceng.llm08.supabase-match-documents-rpc
- seceng.postgres-ai-table-without-rls
- seceng.supabase.rls-not-enabled-for-ai-table
- seceng.github-ai-pr-bot-commits-without-approval
- seceng.gitlab-ai-generated-commit-no-approval
- seceng.llm06.dangerous-tool-without-obvious-approval
- seceng.llm06.tool-call-without-approval-gate
CWE-798 Use of Hard-coded Credentials
Section titled “CWE-798 Use of Hard-coded Credentials”56 rules · confidence P0 3, P1 21, P2 32 · severity error 18, warning 37, info 1
Edges inspected: reaches 3 · calls 1 · reads 13 · sends 2 · grants 37
30 of 56 rules, surest first
- seceng.agentcfg.mcp-env-inline-secret-known-prefix
- seceng.mcp.py.oauth-client-secret-hardcoded
- seceng.mcp.ts.oauth-client-secret-hardcoded
- keryn.keyliteral.anthropic
- keryn.keyliteral.elevenlabs
- keryn.keyliteral.fireworks
- keryn.keyliteral.google-ai-studio
- keryn.keyliteral.huggingface
- keryn.keyliteral.langfuse
- keryn.keyliteral.langsmith
- keryn.keyliteral.openai
- keryn.keyliteral.openrouter
- keryn.keyliteral.perplexity
- keryn.keyliteral.replicate
- keryn.keyliteral.voyage
- keryn.keyliteral.xai
- seceng.agentcfg.mcp-header-inline-credential
- vulnfix-ai-ai_hardcoded_provider_key
- keryn.rails.ruby.basic-auth-hardcoded-password
- keryn.rails.ruby.secret-key-literal
- keryn.rails.yaml.secret-key-base-literal
- seceng.container-ai-secret-env-direct
- seceng.edge.vercel-edge-config-secret-or-prompt
- seceng.tauri-ai-prompt-or-key-in-frontend
- keryn.keyliteral.assigned-literal-provider-key
- keryn.keyliteral.groq
- keryn.keyliteral.pinecone
- keryn.keyliteral.tavily
- seceng.agentcfg.mcp-env-inline-secret-long-value
- seceng.bundle-hardcoded-ai-key
CWE-918 Server-Side Request Forgery
Section titled “CWE-918 Server-Side Request Forgery”56 rules · confidence P0 1, P1 41, P2 14 · severity error 26, warning 30, info 0
Edges inspected: reaches 41 · calls 9 · writes 1 · grants 5
30 of 56 rules, surest first
- keryn.skillcred.cloud-metadata-endpoint-access
- keryn.browseragent.model-output-to-page-navigation-js
- keryn.browseragent.model-output-to-page-navigation-py
- keryn.general.php.ssrf-request-to-http-client
- seceng.asi07.a2a-push-notification-url-not-allowlisted-python
- seceng.asi07.a2a-push-notification-url-not-allowlisted-ts
- seceng.mcp.csharp.tool-http-client-url
- seceng.mcp.go.tool-http-request-ssrf
- seceng.mcp.java.tool-url-fetch
- seceng.mcp.kotlin.tool-url-fetch
- seceng.mcp.rust.tool-reqwest-url
- seceng.modelsc.hf-inference-client-user-model
- seceng.ragloader.ingest-pipeline-fetch-user-url
- seceng.ragloader.langchain-web-loader-ssrf
- seceng.ragloader.langchainjs-recursive-url-loader-ssrf
- seceng.ragloader.langchainjs-web-loader-ssrf
- seceng.ragloader.llamaindex-web-reader-ssrf
- seceng.ragloader.playwright-url-loader-ssrf
- seceng.ragloader.recursive-url-loader-ssrf
- seceng.ragloader.unstructured-url-loader-ssrf
- vulnfix-ai-ai_lc_allow_dangerous_requests
- vulnfix-ai-ai_lc_requests_get_tool
- vulnfix-ai-ai_lc_requests_post_tool
- vulnfix-ai-ai_lc_requests_wrapper
- vulnfix-ai-ai_mcp_param_ssrf
- vulnfix-ai-ai_mcp_tool_ssrf
- keryn.agents2026.openai-agents-function-tool-network-or-file
- keryn.agents2026.openai-agents-js-tool-execute-file-or-network
- keryn.gaps.parameter-url-fetched
- keryn.general.csharp.ssrf-request-to-http-client
CWE-250 Execution with Unnecessary Privileges
Section titled “CWE-250 Execution with Unnecessary Privileges”55 rules · confidence P0 5, P1 40, P2 10 · severity error 23, warning 29, info 3
Edges inspected: calls 11 · writes 1 · grants 42 · depends_on 1
30 of 55 rules, surest first
- keryn.skillcred.container-escape-from-skill
- seceng.agentcfg.mcp-docker-privileged
- seceng.agentcfg.mcp-docker-sensitive-mount
- seceng.asi10.agent-sandbox-host-escape-python
- seceng.asi10.agent-sandbox-host-escape-ts
- keryn.skillexec.download-then-sudo-run
- keryn.skillexec.sudoers-nopasswd
- seceng.agentcfg.allow-bash-unrestricted
- seceng.agentcfg.allow-global-wildcard
- seceng.agentcfg.allow-risky-bash-prefix
- seceng.agentcfg.default-mode-bypass-permissions
- seceng.asi03.agent-sandbox-runs-as-root-python
- seceng.asi03.agent-sandbox-runs-as-root-ts
- seceng.asi03.tool-uses-static-privileged-credential-ts
- seceng.ciagent.agent-allowed-tools-bash-unrestricted
- seceng.ciagent.agent-id-token-write-on-untrusted-trigger
- seceng.ciagent.agent-skip-permissions-flag
- seceng.ciagent.agent-step-on-pull-request-target
- seceng.ciagent.agent-write-permissions-on-untrusted-trigger
- seceng.iacai.agent-tool-workload-host-network
- seceng.iacai.cfn-bedrock-iam-wildcard-action-resource
- seceng.iacai.cfn-bedrock-iam-wildcard-action-resource-json
- seceng.iacai.gpu-ai-container-privileged
- keryn.agentcfg.hook-command-unconditional-allow
- keryn.agentclient.codex-sandbox-full-access
- keryn.agentclient.skill-allowed-tools-unrestricted
- keryn.agenthooks.copilot-hook-uses-sudo
- keryn.agenthooks.cursor-gate-hook-always-allows
- keryn.browseragent.browser-agent-attaches-to-live-browser-js
- keryn.browseragent.browser-agent-attaches-to-live-browser-py
CWE-306 Missing Authentication for Critical Function
Section titled “CWE-306 Missing Authentication for Critical Function”52 rules · confidence P0 2, P1 24, P2 26 · severity error 1, warning 41, info 10
Edges inspected: reaches 14 · calls 3 · sends 1 · grants 30 · trusts 4
30 of 52 rules, surest first
- seceng.eval.public-dashboard-no-auth
- seceng.workflow.webhook-to-credentialed-action
- seceng.iacai.k8s-ai-service-exposed-loadbalancer-nodeport
- keryn.browseragent.browser-agent-attaches-to-live-browser-js
- keryn.browseragent.browser-agent-attaches-to-live-browser-py
- keryn.browseragent.mcp-browser-server-attached-to-user-browser
- seceng.asi07.a2a-agent-card-without-security-schemes-python
- seceng.asi07.a2a-agent-card-without-security-schemes-ts
- seceng.cloudai.vertex-public-endpoint-or-logging
- seceng.gateway-localai-uploads-enabled-public
- seceng.iacai.vllm-container-no-api-key
- seceng.mcp.go.go-sdk-http-bind-all
- seceng.mcp.go.mark3labs-http-bind-all
- seceng.mcp.py.network-transport-bind-all
- seceng.mcp.py.uvicorn-mcp-app-bind-all
- seceng.mcp.ts.listen-bind-all
- vulnfix-ai-ai_mcp_bind_all
- vulnfix-ai-ai_mcp_http_transport
- vulnfix-ai-ai_mcp_sse_transport
- vulnfix-ai-ai_route_no_auth_dependency
- seceng.cloudai.azure-openai-public-network-access
- seceng.container-compose-ai-port-public-no-auth
- seceng.container.ai-service-loadbalancer-no-auth-hint
- seceng.deploy.fly-ai-service-public-no-auth
- seceng.deploy.render-railway-ai-public-no-auth
- seceng.redis-public-cache-no-auth
- keryn.fastapi.llm-route-without-auth
- keryn.flask.llm-route-without-auth
- seceng.agentcfg.mcp-remote-no-authorization
- seceng.asi07.a2a-jsonrpc-endpoint-without-auth-python
CWE-502 Deserialization of Untrusted Data
Section titled “CWE-502 Deserialization of Untrusted Data”49 rules · confidence P0 8, P1 33, P2 8 · severity error 28, warning 19, info 2
Edges inspected: reaches 1 · calls 3 · reads 1 · grants 1 · depends_on 43
30 of 49 rules, surest first
- keryn.general.ruby.unsafe-deserialization-request
- keryn.modelload.cloudpickle-or-dill-loads-py
- keryn.modelload.hf-hub-download-result-unpickled-py
- keryn.modelload.keras-load-model-safe-mode-disabled-py
- keryn.modelload.numpy-load-allow-pickle-py
- keryn.modelload.torch-hub-load-state-dict-nonliteral-url-py
- keryn.modelload.yaml-load-unsafe-loader-py
- seceng.llm05.taint.model-output-to-deserialization-python
- keryn.general.csharp.insecure-deserialization-javascriptserializer-typeresolver
- keryn.general.php.unserialize-untrusted-input
- keryn.modelload.pandas-read-pickle-py
- keryn.modelload.request-path-to-model-loader-python
- keryn.modelload.torch-package-importer-py
- keryn.vfgap.insecure-formatter-deserialization-cs
- keryn.vfgap.json-net-typenamehandling-cs
- keryn.vfgap.jsonpickle-decode-py
- keryn.vfgap.node-serialize-unserialize-js
- keryn.vfgap.snakeyaml-unsafe-load-java
- keryn.vfgap.xmldecoder-java
- keryn.vfgap.xstream-from-xml-java
- seceng.modelsc.transformers-user-controlled-model-id
- vulnfix-ai-ai_js_lc_allow_dangerous
- vulnfix-ai-ai_keras_load_model
- vulnfix-ai-ai_langchain_load_secrets_env
- vulnfix-ai-ai_lc_allow_dangerous_deser
- vulnfix-ai-ai_mcp_param_deser
- vulnfix-ai-ai_mlflow_load_model
- vulnfix-ai-ai_torch_load_unsafe
- keryn.gaps.bytes-from-network-deserialized
- keryn.general.python.marshal-load
CWE-494 Download of Code Without Integrity Check
Section titled “CWE-494 Download of Code Without Integrity Check”47 rules · confidence P0 5, P1 30, P2 12 · severity error 18, warning 26, info 3
Edges inspected: calls 7 · sends 1 · grants 3 · depends_on 36
30 of 47 rules, surest first
- keryn.agenthooks.copilot-hook-pipes-download-to-shell
- keryn.agenthooks.cursor-hook-pipes-download-to-shell
- keryn.agenthooks.gemini-hook-pipes-download-to-shell
- keryn.agenthooks.windsurf-hook-pipes-download-to-shell
- seceng.agentcfg.hook-pipes-to-shell
- keryn.skillexec.download-chmod-run
- keryn.skillexec.download-then-sudo-run
- keryn.skillexec.download-to-temp-exec
- keryn.skillexec.fetch-to-interpreter
- keryn.skillexec.interpreter-oneliner-fetch-exec
- keryn.skillexec.macos-quarantine-bypass
- keryn.skillexec.staging-host-exec
- seceng.agentcfg.mcp-npx-unpinned-package
- seceng.agentcfg.mcp-uvx-unpinned-package
- seceng.asi01.agent-instructions-loaded-at-runtime-python
- seceng.asi01.agent-instructions-loaded-at-runtime-ts
- vulnfix-ai-ai_llamaindex_download_loader
- vulnfix-ai-ai_torch_hub_load
- keryn.agenthooks.copilot-hook-edits-workflows
- keryn.gaps.bytes-from-network-deserialized
- keryn.modelload.torch-hub-load-state-dict-literal-url-unhardened-py
- keryn.skillpersist.raw-ip-or-risky-tld-fetch
- seceng.agentrules.pip-install-from-url
- seceng.agentrules.pipe-to-shell
- seceng.github.actions-unpinned-third-party-action
- seceng.llm03.dynamic-model-download
- seceng.llm03.remote-model-url-without-integrity
- seceng.modelsc.dockerfile-npx-npm-global-unpinned-mcp
- seceng.modelsc.hf-revision-mutable-ref
- seceng.modelsc.mcp-stdio-npx-unpinned-python
CWE-89 SQL Injection
Section titled “CWE-89 SQL Injection”43 rules · confidence P0 7, P1 22, P2 14 · severity error 21, warning 22, info 0
Edges inspected: calls 43
30 of 43 rules, surest first
- keryn.rustagent.model-output-to-sql-rs
- seceng.llm05.taint.model-output-to-sql-js
- seceng.llm05.taint.model-output-to-sql-python
- seceng.mcp.csharp.tool-sql-command-interpolated
- seceng.mcp.go.tool-sql-string-built
- seceng.mcp.java.tool-jdbc-statement-concat
- seceng.mcp.rust.tool-sqlx-query-format
- keryn.general.csharp.sql-injection-request-to-query
- keryn.general.go.sql-injection-request-to-query
- keryn.general.java.sql-injection-request-to-query
- keryn.general.js.sql-injection-request-to-query
- keryn.general.kotlin.sql-injection-request-to-query
- keryn.general.php.sql-injection-request-to-query
- keryn.general.python.sql-injection-request-to-query
- keryn.general.ruby.sql-injection-request-multiline-literal
- keryn.general.ruby.sql-injection-request-to-query
- keryn.general.swift.sql-injection-request-to-query
- vulnfix-ai-ai_lc_sql_database_chain
- vulnfix-ai-ai_mcp_param_sql
- vulnfix-ai-ai_mcp_tool_sql
- vulnfix-ai-ai_taint_sql
- keryn.agents2026.pydantic-ai-tool-argument-to-sql-or-file
- keryn.general.python.django-extra-dynamic-sql
- keryn.general.python.django-rawsql-dynamic
- keryn.rails.cve.dynamic-finder-mysql-cve-2012-6496
- keryn.rails.deps.pg-sqli-cve-2014-0080
- seceng.llm05.prisma-query-raw-unsafe
- seceng.llm05.python-sql-string-format
- seceng.postgres-ai-function-dynamic-sql
- keryn.agents2026.mastra-tool-execute-file-db-network
CWE-532 Insertion of Sensitive Information into Log File
Section titled “CWE-532 Insertion of Sensitive Information into Log File”40 rules · confidence P0 1, P1 20, P2 19 · severity error 6, warning 27, info 7
Edges inspected: reaches 1 · calls 1 · reads 8 · writes 1 · sends 25 · grants 4
30 of 40 rules, surest first
- seceng.eval.mlflow-log-raw-prompt
- keryn.egress.cs.secret-to-side-channel
- keryn.egress.go.secret-to-side-channel
- keryn.egress.java.secret-to-side-channel
- keryn.egress.php.secret-to-side-channel
- keryn.egress.py.secret-to-side-channel
- keryn.egress.ts.secret-to-side-channel
- keryn.skillcred.environment-dump-out
- keryn.skillcred.shell-history-read
- seceng.cloudflare-ai-gateway-logs-prompts
- seceng.csharp-prompt-or-completion-logged
- seceng.eval.phoenix-trace-raw-data
- seceng.gateway.litellm-prompt-logging-enabled
- seceng.gateway.log-contains-api-key
- seceng.java-prompt-or-completion-logged
- seceng.laravel-queue-ai-job-logs-prompt
- seceng.cloudai.bedrock-prompt-logging-or-tracing
- seceng.eval.langfuse-raw-input-output-logging
- seceng.llm07.system-prompt-in-logging
- seceng.postgres-ai-audit-log-raw-prompt
- seceng.redaction-langsmith-langfuse-capture-raw-prompts
- keryn.agents2026.pydantic-ai-instrument-prompts-to-exporter
- keryn.gaps.prompt-or-messages-logged-raw
- keryn.mcp.py.tool-returns-credential-like-keys
- keryn.mcp.ts.tool-returns-credential-like-keys
- seceng.deno.edge-secret-logged
- seceng.deploy-env-secret-logged-near-ai
- seceng.forensics-sensitive-ai-audit-raw
- seceng.gateway.upstream-key-in-response
- seceng.github-token-to-llm-prompt-or-log
CWE-1427 Improper Neutralization of Input Used for LLM Prompting
Section titled “CWE-1427 Improper Neutralization of Input Used for LLM Prompting”37 rules · confidence P0 0, P1 25, P2 12 · severity error 9, warning 28, info 0
Edges inspected: reaches 19 · calls 5 · writes 6 · grants 3 · trusts 4
30 of 37 rules, surest first
- seceng.asi01.system-prompt-from-retrieved-or-tool-content-python
- seceng.asi01.system-prompt-from-retrieved-or-tool-content-ts
- seceng.asi06.memory-read-into-system-prompt-python
- seceng.asi06.memory-read-into-system-prompt-ts
- seceng.asi07.peer-agent-message-as-instructions-python
- seceng.asi07.peer-agent-message-as-instructions-ts
- seceng.mcp.py.tool-description-instructions
- seceng.mcp.py.tool-docstring-instructions
- seceng.mcp.ts.tool-description-instructions
- keryn.browseragent.page-content-in-system-prompt-js
- keryn.browseragent.page-content-in-system-prompt-py
- keryn.llm01.vendor-cohere-system-prompt-taint-js
- keryn.llm01.vendor-cohere-system-prompt-taint-py
- keryn.llm01.vendor-ollama-system-prompt-taint-js
- keryn.llm01.vendor-ollama-system-prompt-taint-py
- keryn.llm01.vendor-openai-assistants-system-prompt-taint-js
- keryn.llm01.vendor-openai-assistants-system-prompt-taint-py
- seceng.iacai.bedrock-guardrail-prompt-attack-none
- seceng.iacai.cfn-bedrock-guardrail-prompt-attack-none
- seceng.iacai.cfn-bedrock-guardrail-prompt-attack-none-json
- seceng.llm01.taint.model-output-to-system-prompt-js
- seceng.llm01.taint.request-to-system-prompt-js
- seceng.llm01.taint.request-to-system-prompt-python
- seceng.llm01.taint.retrieved-content-to-system-prompt-js
- seceng.llm01.taint.retrieved-content-to-system-prompt-python
- keryn.agents2026.openai-agents-web-search-with-write-tools
- keryn.mcp.py.tool-returns-fetched-content-unwrapped
- keryn.mcp.ts.tool-returns-fetched-content-unwrapped
- keryn.skilldirective.cross-agent-addressing
- keryn.skilldirective.document-poisoning-directive
CWE-770 Allocation of Resources Without Limits or Throttling
Section titled “CWE-770 Allocation of Resources Without Limits or Throttling”36 rules · confidence P0 1, P1 13, P2 22 · severity error 5, warning 20, info 11
Edges inspected: reaches 3 · calls 6 · writes 3 · sends 1 · grants 23
30 of 36 rules, surest first
- keryn.skillpersist.fork-bomb
- seceng.asi08.unbounded-agent-fanout-from-model-output-python
- seceng.asi08.unbounded-agent-fanout-from-model-output-ts
- seceng.modelsc.ollama-pull-user-model
- seceng.modelsc.ollama-pull-user-model-ts
- keryn.gaps.unbounded-model-loop
- keryn.skillpersist.disk-or-memory-fill
- seceng.agentframework.crewai-process-without-rate-or-step-limit
- seceng.deploy-ai-service-no-rate-limit-hint
- seceng.edge-ai-route-no-rate-limit
- seceng.gateway-no-per-user-budget
- seceng.llm06.agent-broad-tools-without-approval
- seceng.loop.streaming-without-cancel-or-timeout
- seceng.neon-data-api-ai-route-no-rate-limit
- keryn.agents2026.genkit-generate-tools-no-maxturns
- keryn.agents2026.mastra-agent-call-write-tools-no-maxsteps
- keryn.agents2026.openai-agents-js-run-no-max-turns
- keryn.agents2026.openai-agents-runner-no-max-turns
- keryn.agents2026.pydantic-ai-excessive-retries
- keryn.agents2026.pydantic-ai-run-with-tools-no-usage-limits
- keryn.agents2026.vercelai-v5-stopwhen-excessive-steps
- keryn.fastapi.llm-route-without-rate-limit
- keryn.fastapi.upload-to-llm-without-size-check
- seceng.asi10.scheduled-agent-without-budget-python
- seceng.asi10.scheduled-agent-without-budget-ts
- seceng.deploy-public-ollama-localai-vllm
- seceng.loop.agent-loop-without-max-steps-python
- seceng.loop.agent-loop-without-max-steps-ts
- seceng.loop.langchain-agent-no-max-iterations
- seceng.loop.langgraph-recursion-limit-missing
CWE-345 Insufficient Verification of Data Authenticity
Section titled “CWE-345 Insufficient Verification of Data Authenticity”32 rules · confidence P0 7, P1 15, P2 10 · severity error 21, warning 10, info 1
Edges inspected: reaches 2 · calls 4 · writes 5 · grants 5 · depends_on 3 · trusts 13
30 of 32 rules, surest first
- keryn.cryptoclaim.cs.jwt-verification-disabled
- keryn.cryptoclaim.go.jwt-verification-disabled
- keryn.cryptoclaim.java.jwt-verification-disabled
- keryn.cryptoclaim.kotlin.jwt-verification-disabled
- keryn.cryptoclaim.php.jwt-verification-disabled
- keryn.cryptoclaim.py.jwt-verification-disabled
- keryn.cryptoclaim.ts.jwt-verification-disabled
- keryn.config.jwt-algorithm-none
- seceng.asi01.agent-instructions-loaded-at-runtime-python
- seceng.asi01.agent-instructions-loaded-at-runtime-ts
- seceng.asi01.system-prompt-from-retrieved-or-tool-content-python
- seceng.asi01.system-prompt-from-retrieved-or-tool-content-ts
- seceng.asi06.langgraph-langchain-memory-write-from-untrusted-python
- seceng.asi06.memory-read-into-system-prompt-python
- seceng.asi06.memory-read-into-system-prompt-ts
- seceng.asi07.peer-agent-message-as-instructions-python
- seceng.asi07.peer-agent-message-as-instructions-ts
- seceng.mcp.py.elicitation-auto-accept
- seceng.mcp.py.token-verifier-no-audience-check
- seceng.mcp.ts.elicitation-auto-accept
- seceng.mcp.ts.verify-access-token-no-audience
- seceng.promptsc.prompt-template-from-database-without-version
- keryn.mcp.csharp.sampling-request-no-approval
- keryn.mcp.go.sampling-request-no-approval
- keryn.mcp.java.sampling-request-no-approval
- seceng.asi06.mem0-zep-letta-write-from-untrusted-python
- seceng.asi06.memory-write-from-untrusted-ts
- seceng.asi06.vector-upsert-untrusted-without-source-python
- seceng.asi06.vector-upsert-untrusted-without-source-ts
- seceng.mcp.py.sampling-callback-no-user-approval
CWE-312 Cleartext Storage of Sensitive Information
Section titled “CWE-312 Cleartext Storage of Sensitive Information”29 rules · confidence P0 4, P1 12, P2 13 · severity error 2, warning 15, info 12
Edges inspected: reads 1 · writes 8 · sends 9 · grants 11
29 rules
- seceng.agentcfg.mcp-env-inline-secret-known-prefix
- seceng.eval.mlflow-log-raw-prompt
- seceng.eval.public-dashboard-no-auth
- seceng.eval.weave-trace-without-redaction
- seceng.agentcfg.mcp-header-inline-credential
- keryn.skillcred.service-credential-config-read
- keryn.skillcred.vcs-package-token-read
- seceng.deno.edge-kv-ai-memory-without-tenant-prefix
- seceng.eval.phoenix-trace-raw-data
- seceng.gateway.log-contains-api-key
- seceng.wordpress-transcript-or-prompt-option-storage
- seceng.agentframework.autogen-shared-workdir-or-history
- seceng.agentframework.crewai-memory-without-tenant-scope
- seceng.eval.langfuse-raw-input-output-logging
- seceng.langgraph-state-persists-sensitive-context
- seceng.redis-public-cache-no-auth
- keryn.skillcred.cloud-credential-file-read
- seceng.agentcfg.mcp-env-inline-secret-long-value
- seceng.chromeext-ai-key-in-storage
- seceng.gateway.python-key-in-response
- seceng.gateway.upstream-key-in-response
- seceng.kotlin-transcript-sharedpreferences-storage
- seceng.llm07.frontend-prompt-storage
- seceng.react.chat-history-local-storage
- seceng.swift-transcript-userdefaults-storage
- seceng.langgraph-checkpoint-shared-or-unscoped
- seceng.redis.ai-cache-without-tenant-prefix
- seceng.redis.ai-cache-without-ttl
- seceng.redis.prompt-cache-without-redaction-or-encryption
CWE-73 External Control of File Name or Path
Section titled “CWE-73 External Control of File Name or Path”28 rules · confidence P0 0, P1 18, P2 10 · severity error 12, warning 14, info 2
Edges inspected: reaches 2 · calls 9 · reads 8 · writes 9
28 rules
- keryn.rustagent.model-output-to-fs-path-rs
- seceng.mcp.csharp.tool-file-write
- seceng.mcp.go.tool-file-delete
- seceng.mcp.go.tool-file-write
- seceng.mcp.java.tool-files-write
- seceng.mcp.rust.tool-fs-remove
- seceng.mcp.rust.tool-fs-write
- seceng.modelsc.model-file-user-controlled-path
- seceng.ragloader.directory-loader-path-traversal
- seceng.ragloader.langchain-file-loader-path-traversal
- seceng.ragloader.langchainjs-file-loader-path-traversal
- seceng.ragloader.simple-directory-reader-path-traversal
- keryn.general.ruby.process-stdin-from-request
- keryn.rustagent.model-output-to-desktop-open-rs
- seceng.llm05.taint.model-output-to-file-path-python
- seceng.llm05.taint.model-output-to-filesystem-operation-python
- seceng.llm05.taint.model-output-to-fs-path-js
- seceng.mcp.v3.tool-file-write-from-args-py
- keryn.agents2026.google-adk-tool-function-shell-or-file-write
- keryn.general.ruby.render-instance-variable-from-request
- seceng.csharp.semantic-kernel-file-path-from-args
- seceng.electron-ipc-file-path-from-args
- seceng.java-ai-tool-file-path-from-args
- seceng.llm06.model-controlled-file-path
- seceng.mcp.v3.tool-file-write-from-args
- seceng.tauri.command-file-write-from-args
- keryn.general.js.fs-destination-not-constant
- keryn.general.js.fs-path-not-constant
CWE-506 Embedded Malicious Code
Section titled “CWE-506 Embedded Malicious Code”26 rules · confidence P0 14, P1 10, P2 2 · severity error 23, warning 3, info 0
Edges inspected: calls 9 · writes 5 · sends 11 · depends_on 1
26 rules
- keryn.agenthooks.copilot-hook-reverse-shell
- keryn.agenthooks.cursor-hook-reverse-shell
- keryn.agenthooks.gemini-hook-reverse-shell
- keryn.agenthooks.windsurf-hook-reverse-shell
- keryn.skillexec.reverse-shell-bash-devtcp
- keryn.skillexec.reverse-shell-netcat
- keryn.skillexec.reverse-shell-python
- keryn.skillexec.reverse-shell-script-langs
- keryn.skillexec.reverse-shell-socat
- keryn.skillpersist.crontab-write
- keryn.skillpersist.launchd-agent-install
- keryn.skillpersist.shell-rc-download-exec
- keryn.skillpersist.systemd-unit-enable
- keryn.skillpersist.windows-autorun-or-task
- keryn.agenthooks.copilot-hook-base64-exec
- keryn.agenthooks.cursor-hook-base64-exec
- keryn.agenthooks.gemini-hook-base64-exec
- keryn.agenthooks.windsurf-hook-base64-exec
- keryn.skillexec.decrypt-pipe-to-interpreter
- keryn.skillexec.encrypted-archive-exec
- keryn.skillexec.reverse-shell-powershell
- keryn.skillexec.staging-host-exec
- seceng.agentrules.base64-decode-exec
- keryn.skillpersist.git-hook-install
- keryn.agenthooks.kiro-hook-base64-exec
- keryn.agenthooks.kiro-hook-reverse-shell
CWE-346 Origin Validation Error
Section titled “CWE-346 Origin Validation Error”24 rules · confidence P0 2, P1 12, P2 10 · severity error 8, warning 15, info 1
Edges inspected: reaches 13 · grants 11
24 rules
- seceng.mcp.py.dns-rebinding-protection-disabled
- seceng.mcp.ts.dns-rebinding-protection-disabled
- keryn.browseragent.browser-launch-disable-web-security-js
- keryn.browseragent.browser-launch-disable-web-security-py
- keryn.fastapi.cors-wildcard-with-credentials
- keryn.rustagent.tauri-capability-granted-to-remote-urls
- keryn.rustagent.tauri-config-remote-domain-ipc
- seceng.mcp.ts.cors-wildcard-mcp-server
- seceng.mcp.go.go-sdk-http-bind-all
- seceng.mcp.go.mark3labs-http-bind-all
- seceng.mcp.py.network-transport-bind-all
- seceng.mcp.py.uvicorn-mcp-app-bind-all
- seceng.mcp.ts.listen-bind-all
- seceng.bundle-chrome-message-handler-privileged-api
- seceng.chromeext-externally-connectable-broad
- seceng.chromeext-message-handler-missing-sender-check
- seceng.cloudflare-worker-cors-wildcard-ai
- seceng.deno.edge-cors-wildcard-ai-endpoint
- seceng.manifest-externally-connectable-broad-ai-extension
- seceng.mcp.py.sse-transport-no-security-settings
- seceng.mcp.py.streamable-manager-no-security-settings
- seceng.mcp.ts.sse-transport-no-origin-validation
- seceng.mcp.ts.streamable-http-transport-no-dns-protection
- seceng.supabase-edge-cors-wildcard-with-service-role
CWE-522 Insufficiently Protected Credentials
Section titled “CWE-522 Insufficiently Protected Credentials”23 rules · confidence P0 7, P1 12, P2 4 · severity error 12, warning 11, info 0
Edges inspected: sends 1 · grants 22
23 rules
- keryn.skillcred.browser-credential-store-read
- keryn.skillcred.cloud-metadata-endpoint-access
- keryn.skillcred.macos-keychain-secret-read
- keryn.skillcred.os-credential-database-access
- keryn.skillcred.secret-manager-or-keyring-export
- keryn.skillcred.ssh-key-material-read
- seceng.agentrules.secrets-to-url-command
- seceng.ciagent.agent-step-env-secrets
- seceng.mcp.csharp.token-passthrough
- seceng.mcp.go.token-passthrough
- seceng.mcp.py.token-passthrough
- seceng.mcp.ts.token-passthrough
- keryn.browseragent.browser-agent-uses-personal-profile-js
- keryn.browseragent.browser-agent-uses-personal-profile-py
- keryn.skillcred.cloud-secret-env-or-token-print
- keryn.skillcred.dotenv-file-read-or-upload
- keryn.skillcred.filesystem-secret-sweep
- keryn.skillcred.service-credential-config-read
- keryn.skillcred.vcs-package-token-read
- keryn.skillcred.cloud-credential-file-read
- keryn.skillcred.crypto-wallet-file-access
- keryn.skillcred.other-agent-credential-read
- seceng.llm03.huggingface-hardcoded-token-login
CWE-77 Command Injection
Section titled “CWE-77 Command Injection”22 rules · confidence P0 0, P1 14, P2 8 · severity error 1, warning 21, info 0
Edges inspected: reaches 15 · calls 2 · grants 5
22 rules
- seceng.ciagent.agent-prompt-interpolates-event-text
- keryn.gaps.retrieved-text-in-system-role
- keryn.llm01.vendor-cohere-system-prompt-taint-js
- keryn.llm01.vendor-cohere-system-prompt-taint-py
- keryn.llm01.vendor-ollama-system-prompt-taint-js
- keryn.llm01.vendor-ollama-system-prompt-taint-py
- keryn.llm01.vendor-openai-assistants-system-prompt-taint-js
- keryn.llm01.vendor-openai-assistants-system-prompt-taint-py
- seceng.agentrules.conceal-from-user
- seceng.agentrules.hidden-html-comment-instructions
- seceng.agentrules.instruction-override-phrase
- seceng.llm01.taint.model-output-to-system-prompt-js
- seceng.llm01.taint.request-to-system-prompt-js
- seceng.llm01.taint.retrieved-content-to-system-prompt-js
- keryn.agents2026.openai-agents-js-powerful-tool-no-input-guardrails
- keryn.agents2026.openai-agents-powerful-tool-no-input-guardrails
- keryn.gaps.instructions-concatenated-with-input
- seceng.agentrules.mandatory-fetch-or-exec
- seceng.llm01.langchain-sink-injection
- seceng.llm01.llamaindex-sink-injection
- seceng.ragloader.retrieved-content-in-system-prompt
- seceng.ragloader.retrieved-content-in-system-prompt-ts
CWE-116 Improper Encoding or Escaping of Output
Section titled “CWE-116 Improper Encoding or Escaping of Output”21 rules · confidence P0 2, P1 14, P2 5 · severity error 11, warning 6, info 4
Edges inspected: reaches 6 · calls 7 · sends 3 · grants 5
21 rules
- seceng.mcp.py.tool-description-hidden-unicode
- seceng.mcp.ts.tool-description-hidden-unicode
- keryn.skillexec.base64-exec-script-langs
- keryn.skillexec.powershell-base64-iex
- keryn.skillexec.powershell-encodedcommand
- seceng.agentrules.bidi-override-character
- seceng.agentrules.invisible-unicode-character
- seceng.aiconfig.bidi-override-character
- seceng.aiconfig.escaped-invisible-unicode
- seceng.aiconfig.invisible-unicode-character
- seceng.llm01.prompt-injection-taint
- keryn.general.js.template-escape-markup-disabled
- seceng.llm01.js-ts-concat-prompt
- seceng.llm01.prompt-injection-taint-js
- seceng.llm01.trusted-dev-prompt-suppression
- seceng.vercelai-stream-without-moderation-buffer
- seceng.llm01.openai-assistants-thread-injection
- seceng.llm01.python-concat-prompt
- seceng.llm01.system-user-role-flip-contextual
- seceng.stream.sse-direct-llm-stream-no-filter-ts
- seceng.stream.websocket-direct-llm-output
CWE-915
Section titled “CWE-915”20 rules · confidence P0 0, P1 4, P2 16 · severity error 0, warning 19, info 1
Edges inspected: calls 6 · reads 2 · writes 9 · grants 3
20 rules
- keryn.general.ruby.mass-assignment-permit-all
- keryn.outbound.model-output-to-crm-record-js
- keryn.outbound.model-output-to-crm-record-py
- keryn.rails.ruby.mass-assignment-without-protection
- keryn.general.ruby.mass-assignment-unpermitted-params
- keryn.rails.cve.create-with-cve-2014-3514
- keryn.rails.legacy.model-attr-protected
- keryn.rails.legacy.model-without-attr-accessible
- keryn.rails.ruby.permit-privileged-attribute
- seceng.firebase.ai-output-direct-firestore-write
- seceng.laravel-ai-output-direct-db-write
- seceng.llm06.tool-output-direct-db-write
- seceng.mongoose-ai-output-direct-write
- seceng.orm.prisma-ai-output-direct-write
- seceng.redis-cache-poisoning-from-model-output
- seceng.schema.model-json-parse-without-validation-python
- seceng.schema.model-json-parse-without-validation-ts
- seceng.supabase-ai-transcript-insert-with-request-user-id
- seceng.vercelai-output-to-database-without-validation
- keryn.general.js.plain-object-dynamic-key-write
CWE-327
Section titled “CWE-327”18 rules · confidence P0 0, P1 16, P2 2 · severity error 9, warning 7, info 2
Edges inspected: sends 7 · depends_on 1 · trusts 9 · none 1
18 rules
- keryn.cryptoclaim.cs.ecb-mode-for-claims
- keryn.cryptoclaim.go.ecb-mode-for-claims
- keryn.cryptoclaim.java.ecb-mode-for-claims
- keryn.cryptoclaim.kotlin.ecb-mode-for-claims
- keryn.cryptoclaim.php.ecb-mode-for-claims
- keryn.cryptoclaim.py.ecb-mode-for-claims
- keryn.cryptoclaim.swift.ecb-mode-for-claims
- keryn.cryptoclaim.ts.ecb-mode-for-claims
- keryn.general.python.ssl-insecure-protocol
- keryn.general.python.ecb-mode
- keryn.general.python.ssl-insecure-protocol-default
- keryn.general.python.ssl-wrap-socket
- keryn.general.python.weak-cipher
- keryn.general.python.weak-crypt-method
- keryn.general.python.weak-hash
- keryn.vfgap.weak-cipher-des-rc4-go
- keryn.general.python.pycrypto-namespace
- keryn.general.python.pyghmi-ipmi
CWE-668 Exposure of Resource to Wrong Sphere
Section titled “CWE-668 Exposure of Resource to Wrong Sphere”18 rules · confidence P0 3, P1 10, P2 5 · severity error 8, warning 10, info 0
Edges inspected: reaches 7 · calls 1 · grants 10
18 rules
- seceng.agentcfg.mcp-docker-sensitive-mount
- seceng.asi10.agent-sandbox-host-escape-python
- seceng.asi10.agent-sandbox-host-escape-ts
- seceng.asi06.memory-shared-user-scope-python
- seceng.asi06.memory-shared-user-scope-ts
- seceng.iacai.agent-tool-workload-host-network
- seceng.iacai.azure-openai-public-network-access-enabled
- vulnfix-ai-ai_gradio_share
- seceng.iacai.azure-openai-no-network-acls
- seceng.iacai.cfn-sagemaker-notebook-direct-internet
- seceng.iacai.cfn-sagemaker-notebook-direct-internet-json
- seceng.iacai.sagemaker-notebook-direct-internet
- seceng.iacai.vertex-ai-endpoint-no-private-network
- keryn.vendor.notebooks-instance-public-ip-tf
- keryn.vendor.workbench-instance-public-ip-tf
- seceng.asi10.code-sandbox-unrestricted-lifetime-or-network-python
- seceng.asi10.code-sandbox-unrestricted-lifetime-or-network-ts
- seceng.iacai.ollama-bind-all-interfaces