agent-brain

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

• Third-party content exposure detected (high risk: 0.85). Outsider-authored free text can enter the LLM context via /download-brain//sync-brain pulling remote MEMORY.md, USER.md, sessions/*.md, and projects/*.md from pCloud, then /load-brain reading those files “full” into the agent’s context; those remote markdown files are not authored by the operating user.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill performs runtime pCloud API calls (e.g. https://api.pcloud.com and the OAuth URL https://my.pcloud.com/oauth2/authorize) to download remote markdown memory files via curl and merge them into the local agent-brain, which are then loaded into the agent's context—meaning remotely hosted content can directly influence agent prompts/behavior.