Intake, deduplication, severity assignment, and prioritization of security findings for bug bounty and pentest engagements. Processes raw findings into actionable triaged items.
Audit third-party dependencies for known vulnerabilities and detect hardcoded secrets, API keys, and credentials in source code, configuration, and git history.
Compile, format, and generate security assessment reports from triaged findings. Produces consistent, professional reports with executive summaries, finding details, remediation guidance, and validation notes.
Develop and execute proof-of-concept exploits to validate suspected vulnerabilities. Confirms exploitability, measures real impact, and produces reproducible evidence for confirmed findings.
Deep reconnaissance review and attack surface mapping. Analyzes gathered recon data to identify exposed services, entry points, technology stacks, and potential attack vectors.
Security and safety assessment for LLM applications, AI agents, and AI-integrated systems. Covers OWASP LLM Top 10 2025, prompt injection, jailbreaking, training data attacks, model extraction, AI plugin abuse, agentic system risks, and AI red teaming methodology.
Security assessment of iOS and Android mobile applications. Covers OWASP Mobile Top 10, reverse engineering, runtime manipulation, data storage, network security, certificate pinning bypass, and platform-specific attack patterns.
Source-level vulnerability hunting. Systematic review of application source code for injection flaws, auth bypasses, logic bugs, cryptographic weaknesses, and unsafe patterns. Operates on code files within the workspace.
Establishes engagement rules, authorization boundaries, and operational guardrails for all security review skills. Must be invoked before any assessment work begins.
Fast-path security audit of files changed since last commit. Runs automatically after every coding task. Scopes exclusively to git diff. Escalates to the full applicable skill if S1 or S2 findings are discovered.
Proactive threat identification using structured frameworks (STRIDE, PASTA). Models data flows, trust boundaries, and threat actors to prioritize security testing and contextualize findings with business risk.