"promptscan" is not a named product in the sources, so these are candidates for scanning LLM apps and agents for prompt injection. Promptfoo 4 and PromptGuard Scan 1 are built for prompt-injection testing in CI, RAXE 2 and Inkog 3 cover agent runtime and static analysis, and Superagent 7 and Bawbel 8 scan agent components.
- 1
- 2
PromptGuard Scanpromptguardscan.spacewebsite read, tools not yet checked
Tests LLM apps for prompt injection, jailbreaks, sensitive data leakage and unsafe tool calls, with a release gate that fails builds on high-risk findings.
- 3
- 4
Inkoginkog.iofound in code, not yet checked
Static analysis of agent code, MCP servers and Agent Skills that traces untrusted text to actions to find prompt injection into tools.
for agents: inkog_scan · inkog_skill_scan · inkog_mcp_scan
- 5
Superagentsuperagent.shfound in code, not yet checked
Scans repositories and agent skills for prompt injection and malicious instructions, and analyzes text for prompt injection via superagent_guard 17.
+2 community tools built for it
for agents: superagent_scan · skill-security · superagent_guard
- 6
Sources
- 1Prompt Injection Scanner for LLM Apps | PromptGuard Scanpromptguardscan.space · read 2026-10-29
- 2RAXE – Runtime Detection & Evidence for AI Agentsraxe.ai · read 2026-10-30
- 3Inkog AI agent security: find prompt injection before it shipsinkog.io · read 2026-10-30
- 4Build Secure AI Applications | Promptfoopromptfoo.dev · read 2026-10-29
- 7Superagent | The autonomous security factorysuperagent.sh · read 2026-10-30
- 8Bawbel: Agentic AI Security Scannerbawbel.io · read 2026-10-30
- 12raxe-ce: declares "scan_prompt" (the record we read)github.com · read 2026-10-24
- 17superagent: declares "superagent_scan" (the record we read)github.com · read 2026-10-24
For agents: the 10 actions behind this answer
· PromptGuard Scan website read, tools not yet checked
PromptGuard Scan tests LLM apps for prompt injection, jailbreaks, sensitive data leakage, unsafe tool calls, and CI/CD release risks.
scan_prompt · raxe-ce found in code, not yet checked
Scan a prompt or text for security threats including
inkog_scan · inkog found in code, not yet checked
Security co-pilot for AI agent development. Scans for prompt injection, infinite loops, token bombing, SQL injection via LLM, and missing guardrails. Supports LangChain, CrewAI, LangGraph, AutoGen, n8n, and 20+ agent frameworks. Use this whenever building, reviewing, or deploying AI agents to catch
run_assertion · promptfoo found in code, not yet checked
· PromptBrake website read, tools not yet checked
Find prompt injection, data leaks, and unsafe AI behavior before you ship. Self-hosted automated testing with evidence, remediation guidance, and CI release gates.
start_codex_security_prompt_only_scan · codex-security found in code, not yet checked
Start or rejoin a Standard or diff Codex Security scan from its owning conversation. Use the returned authoritative scanId and scanDir. Standard and diff scans save progress checkpoints before their final semantic draft; the workbench writes the unsealed canonical artifacts. Complete the same scan o
superagent_scan · superagent found in code, not yet checked
Scan a Git repository for AI agent-targeted attacks like repo poisoning, prompt injection, and malicious instructions using Superagent's security scanning. This tool clones a repository into a secure sandbox and uses AI to analyze its contents for security threats that could exploit AI agents or cod
· Bawbel website read, tools not yet checked
Open-source security scanner for agentic AI components. 51 AVE records, justified suppression, toxic flow detection, MCP conformance scoring.
get_vault_status · Chrome MCP Server (Security Hardened) found in code, not yet checked
Get the status of the credential vault including encryption configuration.
· Panguard website read, tools not yet checked
AI agent security for developers. Scan MCP skills for prompt injection, tool poisoning, and supply chain threats. One command to install. Built on the open ATR standard. Free and open source.
Labels are the instruments' verdict and are never changed by ranking. Agents and robots get this same answer as JSON at /api/v1/find?q=…&answer=1 and through the MCP server.