bug-bounty

The code functions as a targeted IDOR testing harness against HackerOne's GraphQL API, capable of performing numerous privileged mutations using provided account cookies. While not inherently malicious, its capability to alter or disclose information makes it risky if misused or deployed without explicit authorization. The insecure SSL handling and brittle CSRF token extraction further elevate operational risk. A safer, auditable version would constrain mutations, validate inputs, and remove disablement of TLS verification. Overall risk is elevated due to potential destructive actions on real reports.

SUSPICIOUS: the content is internally coherent as a bug bounty triage guide, but it gives an AI agent offensive security validation workflow against live targets and relies on untrusted external content. No credential harvesting, exfiltration, proxy routing, or malicious payload execution is present; risk is mainly from security-research enablement and moderate distribution trust.

SUSPICIOUS due to high-risk offensive security capability for an AI agent, not because of deceptive behavior. The skill is internally consistent and does not show credential harvesting, exfiltration, or dubious install paths, but it materially increases an agent's ability to perform reverse engineering and security analysis on mobile apps.

SUSPICIOUS: the skill is internally consistent as an offensive security reference, but that purpose itself gives an AI agent high-risk exploitation capability against web targets. There is little evidence of credential theft, stealth, or malicious supply-chain behavior; the main concern is enabling active attacks, scans, and exploit chaining on real systems.

The skill is purpose-consistent but high risk: it equips an AI agent with offensive recon, scanning, fuzzing, and secret-hunting capabilities against arbitrary web targets. Install/data flows mostly point to official or well-known OSS tools rather than obvious malware, so this is not confirmed malicious; however, as an AI-agent security/exploit skill with autonomous external actions and broad credential/tool use, it should be classified as SUSPICIOUS/high-risk.

SUSPICIOUS: the skill is internally consistent as a mobile security assessment orchestrator, but it is high-risk because it equips an AI agent to conduct offensive security work and to pivot into broader backend testing through other skills. No direct credential theft, covert exfiltration, or malicious installer behavior is present in this skill itself.

SUSPICIOUS: the skill is internally consistent as a bug-bounty/offensive-security guide, but that stated purpose is itself high risk for an AI agent. It enables recon, exploit testing, prompt-extraction, and indirect-injection workflows against external systems, while also using moderately risky GitHub/raw and unpinned install paths.

SUSPICIOUS: the skill is internally coherent for mobile app dynamic security testing and shows low installer/data-exfiltration concern, but it is a high-risk offensive security skill for an AI agent because it enables MITM, pinning bypass, runtime instrumentation, and follow-on recon against real targets.

This artifact is an explicit, actionable offensive playbook/agent specification that operationalizes chaining vulnerabilities to steal credentials, perform account takeover, escalate privileges, and achieve remote code execution. It provides concrete steps, tooling guidance, and monetization context, and it lacks any safeguards for authorized or defensive use. Treat this as high-risk malicious guidance: do not deploy or run against third-party systems. Restrict to isolated, authorized lab environments for defensive research only.

This fragment is not defensive software; it is a highly actionable exploit-chain playbook containing concrete instructions for conducting multi-stage attacks (OAuth theft→ATO, SSRF to cloud metadata/IAM credential access, XSS→CSRF privilege escalation, and injection/exfiltration patterns). There is no code to analyze for hidden behavior, but the distributed content itself is directly weaponizable and should be treated as extremely high risk and removed/isolated from any legitimate software supply chain.