what-an-email-reveals
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides a shell command template in
SKILL.md(Step 3) for calculating MD5 hashes of email addresses:printf '%s' "jdoe@example.com" | tr 'A-Z' 'a-z' | md5sum. - Evidence: The agent is instructed to "Compute the hash" using this shell pipeline. If an email address containing shell metacharacters (e.g., backticks or
$(...)) is provided and the agent performs raw string interpolation into the command, it could lead to arbitrary command execution on the host environment. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it instructs the agent to process untrusted external data (email headers) which may contain malicious instructions designed to influence the agent's behavior.
- Ingestion points: The agent ingests email headers and message content for analysis in
SKILL.md(Step 5) andreference/header-fields.md. - Boundary markers: The instructions lack explicit boundary markers or directions for the agent to ignore instructions embedded within the header fields (e.g.,
Subject,X-headers, orMessage-ID). - Capability inventory: The skill has the capability to execute shell commands (
bash) and chain to various other investigative tools (e.g.,find-exposed-servers,what-leaked-about-you,secrets-in-git-history). - Sanitization: No sanitization, validation, or escaping of the external header content is described before the agent processes and interprets the data.
- [SAFE]: The skill references several external resources and well-known services (such as Gravatar, Google, and GitHub). These references are consistent with the skill's primary OSINT purpose and follow the trusted service guidelines.
Audit Metadata