email-to-calendar
Fail
Audited by Snyk on Nov 22, 2026
Risk Level: CRITICAL
Full Analysis
CRITICAL E004: Prompt injection detected in skill instructions.
- Potential prompt injection detected (medium risk: 0.30). No clear malicious intent is visible, but the publish step points to a near-lookalike registry domain (auth.clawdhub.com for the "clawhub" tool) and the setup guide directs users to install an unvetted personal fork of gogcli, both sketchy enough to flag under recall-first.
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (medium risk: 0.30). In DIRECT mode, the required workflow scans all unread inbox emails (via
email_search.sh --include-body), and the BOOT.md heartbeat repeats this on a schedule, so any outsider who emails the user's Gmail inbox can place free text the agent ingests without the agent selecting a specific item.
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (medium risk: 0.30). The only inbound code source is the setup instruction in references/gog-commands.md to install gogcli from the individual-owned fork github.com/tonimelisma/gogcli on a feature branch, which is unofficial but attributable and not an unrelated or suspicious source, so it is a common/reputable unofficial dependency and rates medium.
MEDIUM W013: Attempt to modify system services in skill instructions.
- Attempt to modify system services in skill instructions detected (low risk: 0.10). The skill's BOOT.md instructs the agent to edit the persistent HEARTBEAT.md configuration to add scanning and pending-invite sections, which is an operational change to agent startup/heartbeat behavior, but no explicit irreversible operational loss is established (undo, changelog trimming, and tracking removal concern content history rather than needed operational function).
Issues (4)
E004
CRITICALPrompt injection detected in skill instructions.
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
W012
MEDIUMUnverifiable external dependency detected (runtime URL that controls agent).
W013
MEDIUMAttempt to modify system services in skill instructions.
Audit Metadata