android-tools-reviewer

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

• Third-party content exposure detected (high risk: 0.85). Outsider free text can enter the LLM context via the PR narrative: the workflow runs gh pr view {number} --json title,body and then “read[s] the PR description and linked issues,” where title/body are authored by the PR submitter (an outsider) and are ingested as readable text.

MEDIUM W021: Hidden or invisible Unicode characters detected (potential obfuscation or prompt injection).

• Hidden Unicode characters detected (1 type(s) found)