Full Report
A financially motivated threat actor has been linked to the development and distribution of a JavaScript (JS)-based information stealer known as PhantomRaven via the npm package registry. "The developer likely wrote the malware using a large language model (LLM), an assessment made with high confidence based on verbose comments, placeholder code, and statistical token-analysis patterns,"
Analysis Summary
# Threat Actor: PhantomRaven Developer (Self-proclaimed Bug Bounty Hunter)
## Attribution & Identity
* **Actor Identification:** A financially motivated threat actor who poses as a legitimate bug bounty hunter.
* **Aliases/Online Identities:**
* jpdhellonpm1
* jpd15
* jpd12
* jpd13
* npmhell
* npmpackagejpd
* npmtestdharsh
* jpdhackerone11
* packagedharsh
* **Known Associations:** Linked to the development of the "PhantomRaven" information stealer.
## Activity Summary
* **Timeline:** Active since at least November 2022.
* **Recent Campaigns:** In late 2025 and throughout 2026, the actor engaged in a large-scale software supply chain attack using "slopsquatting" and typosquatting on the npm registry. Over 100 malicious packages were uploaded.
* **LLM Integration:** The actor is noted for using Large Language Models (LLMs) to generate malware code, evidenced by verbose comments and specific token patterns.
* **Extortion/Bounty Claims:** In August 2025, the actor claimed Remote Code Execution (RCE) on targets via their own malicious npm packages to solicit rewards from vulnerability disclosure programs.
## Tactics, Techniques & Procedures
* **Supply Chain Attack:** Uploading malicious packages to public registries (npm, PyPI).
* **Slopsquatting/Typosquatting:** Naming packages similarly to popular libraries to deceive developers.
* **Remote Dynamic Dependency (RDD):** Using packages as a "cover" to fetch a second-stage payload from an external server to evade static analysis.
* **Information Stealing:** Scanning environments for Git/npm configurations, SSH/GitHub credentials, and CI/CD secrets.
* **System Fingerprinting:** Collecting public IPs, usernames, and runtime details.
* **LLM-Assisted Development:** Utilizing AI to rapidly generate and iterate proprietary malware code.
* **MITRE ATT&CK IDs (Inferred):**
* T1195.001 (Supply Chain Compromise: Compromise Software Dependencies)
* T1584 (Compromise Infrastructure)
* T1552.004 (Unsecured Credentials: Private Keys/Secrets)
* T1059.003 (Command and Scripting Interpreter: JavaScript)
## Targeting
* **Sectors:** Technology, Retail, and Hospitality.
* **Geography:** Global (targeting users of the npm and PyPI registries).
* **Victims:** Developers and organizations using CI/CD pipelines (GitHub Actions, GitLab CI, Jenkins, CircleCI). The actor claims to have successfully collected bounties from at least nine distinct entities.
## Tools & Infrastructure
* **Malware Families:**
* **PhantomRaven:** A JavaScript-based information stealer.
* Unnamed Python-based stealer (found on PyPI).
* **Infected npm Packages:**
* `transform-jsbi-to-bigint`
* `sort-imports-es6-autofix`
* **Infrastructure:**
* C2/External Servers: Used for hosting Remote Dynamic Dependencies (RDD) and exfiltrating data (Specific defanged URLs not provided in text, but actor used domains associated with the `jpd` and `npm` aliases).
## Implications
* **Strategic Assessment:** This actor represents a shift in "grey-hat" or criminal activity where malware is used to force a bug bounty payout. By creating the vulnerability themselves (via supply chain poisoning), they bypass traditional ethical hacking boundaries.
* **AI Threat Landscape:** High confidence that LLMs are lowering the barrier to entry for creating custom, verbose, and functional malware, allowing single actors to maintain larger quantities of malicious packages.
## Mitigations
* **Package Verification:** Implement strict vetting for npm/PyPI dependencies; use tools to flag newly created or low-reputation packages.
* **Lockfiles & Integrity:** Use `package-lock.json` or `yarn.lock` to ensure sub-dependencies are not unexpectedly swapped for malicious versions.
* **Secret Management:** Mask and rotate CI/CD environment variables; avoid storing plaintext credentials in Git configurations.
* **Egress Filtering:** Monitor and restrict outbound traffic from build servers/developer environments to unknown or suspicious IP addresses/domains.