trannghiach/aurelinth-ctf-agentic-ai

sqli-hunter

CTF web challenge SQL injection hunting and exploitation. Trigger when web-recon has identified injectable endpoints or params, need to confirm, exploit, and extract data including flags from the database.

원문 언어: 영어

web-recon

CTF web challenge reconnaissance. Trigger when starting a new web target, need to enumerate endpoints, identify technologies, discover forms and inputs, or map attack surface before exploitation.

원문 언어: 영어

auth-bypasser

CTF web challenge authentication bypass. Trigger when web-recon has identified login forms, admin panels, JWT tokens, session cookies, or access control mechanisms that need bypass testing.

원문 언어: 영어

crypto-hunter

CTF blackbox cryptographic token hunter. Trigger when web_recon finds JWT tokens, JWE tokens, encrypted cookies, weak session tokens, or exposed public keys. Tests JWT algorithm confusion, weak secrets, JWE public key forge, predictable tokens. Does NOT…

원문 언어: 영어

file-upload-hunter

CTF web challenge file upload bypass and webshell. Trigger when web-recon identifies file upload forms, avatar/attachment endpoints, or import functionality.

원문 언어: 영어

flag-extractor

CTF web challenge final analysis and flag extraction. Trigger after all hunting agents have completed, need to aggregate findings, extract flag from collected data, or escalate remaining leads to get the flag.

원문 언어: 영어

idor-hunter

CTF web challenge IDOR and broken access control. Trigger when web-recon identifies user-specific resources, numeric IDs in URLs, API endpoints, or JWT/session-based access control.

원문 언어: 영어

lfi-hunter

CTF web challenge LFI, path traversal, and RCE via LFI. Trigger when web-recon identifies file parameters, include/require patterns, or path traversal indicators.

원문 언어: 영어

ssti-hunter

CTF web challenge SSTI detection and exploitation. Trigger when web-recon identifies template rendering, Flask/Django/Express apps, or user input reflected in responses suggesting template context.

원문 언어: 영어

xss-hunter

CTF web challenge XSS hunting and exploitation. Trigger when web-recon has identified reflected inputs, user-controlled output, or HTML injection points that need XSS testing.

원문 언어: 영어

access-control-auditor

CTF whitebox IDOR and broken access control auditor. Trigger when vuln_reasoner identifies a missing ownership check or object reference that can be manipulated. Confirms the missing check, identifies the target object ID, exploits directly.

원문 언어: 영어

auth-auditor

CTF whitebox authentication auditor. Trigger when vuln_reasoner identifies an auth bypass, weak JWT, session forgery, or privilege escalation path. Confirms bypass logic, crafts exploit, tests locally, attacks real target.

원문 언어: 영어

code-reader

CTF whitebox challenge source code mapping. Trigger as the first step of any whitebox pipeline to map file structure, identify entry points, trace data flows, and extract auth/database patterns before vulnerability analysis begins.

원문 언어: 영어

crypto-auditor

CTF whitebox crypto auditor. Trigger when vuln_reasoner identifies weak cryptography — predictable token, broken algorithm, key reuse, padding oracle, or custom crypto. Confirms weakness from source, breaks it, forges or decrypts to reach the flag.

원문 언어: 영어

dep-checker

CTF whitebox dependency vulnerability analysis. Trigger after code_reader has completed. Reads dependency files, identifies library versions, checks for known CVEs, and assesses whether any CVE provides a direct exploit path for the current challenge. Outputs…

원문 언어: 영어

deserialization-auditor

CTF whitebox deserialization auditor. Trigger when vuln_reasoner identifies unsafe deserialization — pickle, PyYAML, PHP unserialize, Java ObjectInputStream, or similar. Confirms gadget chain from source, crafts malicious payload, achieves RCE or flag read.

원문 언어: 영어

lfi-auditor

CTF whitebox LFI / path traversal auditor. Trigger when vuln_reasoner identifies a file read or include with user-controlled path. Confirms traversal logic, identifies target file, crafts payload, tests locally, attacks real target.

원문 언어: 영어

race-condition-auditor

CTF whitebox race condition auditor. Trigger when vuln_reasoner identifies a TOCTOU window, concurrent state issue, or check-then-act gap. Confirms the race window from source, builds concurrent exploit, tunes timing locally, attacks real target.

원문 언어: 영어

sqli-auditor

CTF whitebox SQL injection auditor. Trigger when vuln_reasoner identifies a suspected SQLi with a known code location. Confirms exploitability via isolation test, crafts targeted exploit, tests locally, attacks real target.

원문 언어: 영어

ssti-auditor

CTF whitebox SSTI auditor. Trigger when vuln_reasoner identifies a template render call with user-controlled input. Confirms injection context, identifies template engine from imports, crafts RCE or flag-read payload, tests locally, attacks real target.

원문 언어: 영어

upload-auditor

CTF whitebox file upload auditor. Trigger when vuln_reasoner identifies a file upload handler with bypassable validation. Confirms validation logic from source, crafts bypass, uploads malicious file, achieves RCE or flag read.

원문 언어: 영어

vuln-reasoner

CTF whitebox vulnerability analysis. Trigger after code_reader and dep_checker have completed. Reads their structured findings and reasons about exploitability, data flows, and attack paths. Outputs ranked, structured findings for supervisor to dispatch…

원문 언어: 영어

xss-auditor

CTF whitebox XSS auditor. Trigger when vuln_reasoner identifies a suspected XSS with a known template or output location. Confirms rendering context, crafts targeted payload, verifies locally, escalates to flag extraction.

원문 언어: 영어