pentestcode-ai-pentest-agent

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly stores and displays plaintext credentials (e.g., state_update with password, /creds output, programmatic reads of state.creds) and shows example API key environment variables, which requires the LLM/agent to handle and potentially emit secret values verbatim.

---

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.90). The README instructs users to curl a raw install.sh from a GitHub user (s0ld13rr) and pipe it to bash — running an unreviewed shell script from an individual/unknown repo is a common malware distribution pattern and therefore suspicious.

---

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The document describes an agent built to perform offensive operations (credential harvesting, SAM/LSA dumping, credential spraying, lateral movement, pivot tunnels) and includes install patterns that enable supply-chain risk (curl|bash), meaning it contains explicit, high-risk malicious/abusive capabilities that can be used for unauthorized compromise and data theft.

---

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.65). SKILL.md describes runtime parsing of tool outputs (e.g., nmap_parse of scan.xml, nuclei_parse of nuclei.json, gobuster_parse of gobuster.txt) where those files can contain arbitrary, outsider-controlled free text from scanned targets and then be ingested into the agent’s engagement state/LLM context.

---

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The SKILL.md contains install commands that fetch and execute remote code via "curl -fsSL https://raw.githubusercontent.com/s0ld13rr/pentestcode/main/install.sh | bash", which runs external script content at install/runtime.

---

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (medium risk: 0.60). The skill contains explicit privileged installation instructions (curl | bash installers, sudo apt/pacman commands, installing to /usr/local/bin and /opt) and a mode to bypass scope checks, which encourage modifying the host system and using elevated privileges.