Caveat verdict

agent-mail-guard

88
🟢 Trusted
No high-risk patterns surfaced by the deep scan — automated capability review, not behavioral proof.

Email sanitization middleware that explicitly detects and blocks prompt injection, base64 payloads, homoglyphs, and markdown exfiltration; a defensive security tool with no malicious behavior.

⚠ Flagged for review — coarse, uncorroborated signal, not a confirmed exploit. Review the config yourself before installing.

Automated static analysis — not a human review. Caveat flags capabilities, not confirmed intent, and can produce false positives. Disagree with this verdict? Use Dispute below.

0
security
90
transparency
90
maintenance

Findings (10)

Pattern match critical

Pipe to python — executes piped content as Python code

README.md · code · | python3

Pattern match critical

Prompt injection — tries to override agent instructions

README.md · code · Ignore previous instructions

Pattern match critical

Unicode homoglyph detected — uses lookalike characters to evade pattern matching

test_sanitizer.py · prose

Pattern match high

Possible prompt injection — attempts to redefine agent identity

README.md · code · You are now

Pattern match high

subprocess execution — runs system commands from Python

README.md · code · subprocess.run(

Pattern match high

Raw model control tokens — prompt injection via token manipulation

sanitize_core.py · prose · downgraded · [INST]

Pattern match high

Prompt injection — disregard instructions variant

test_sanitizer.py · prose · downgraded · Disregard all previous instructions

Pattern match high

<script> tag in markdown — potential code injection

test_sanitizer.py · prose · downgraded · <script>

Pattern match medium

Instructs covert action — may act without user awareness

README.md · prose · downgraded · silently

Pattern match medium

Fake system prompt — attempts to inject instructions

test_cal_sanitizer.py · prose · downgraded · system: You are

Why the tier is capped

Execution sink present in raw bytes (Hard Floor: class B/D). Final tier capped at Caution — cannot be lifted by any downgrade, example-payload opt-in, or allowlist.

Permissions & capabilities

Requires 1 system binary.

Is this flag fair?

Check another skill Browse the registry Auditing your own skills or configs? Use the API