Caveat verdict
avatar-runtime
Accesses credentials AND makes external network calls
The skill explicitly states it contains no bundled code and instructs the agent to download and execute the avatar-runtime npm package via npx, which is runtime remote code execution; the skill itself acknowledges this and recommends sandboxing first, but it is a genuine concern.
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.
What it does
These are capability combinations: each listed behavior occurs in the skill, but Caveat detects co-occurrence — it does not verify that one flows into another. Read the code to confirm a live chain.
Accesses credentials AND makes external network calls — potential credential theft
LLM02 · ASI03
Permission integrity
network_out
credential_access
Findings (6)
<script> tag in markdown — potential code injection
SKILL.md · code · <script
Pipe to python — executes piped content as Python code
SKILL.md · code · | python3
Possible hardcoded credential
SKILL.md · code · API_KEY="your-key
Pipe-to-python pattern — remote code execution risk
SKILL.md · code · curl -s "$BASE/v1/status" | python
HTTP request to bare IP address — common in malicious payloads
SKILL.md · code · http://127.0.0.1
Instructs covert action — may act without user awareness
references/WEB-EMBEDDING.md · prose · downgraded · silently
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class B/C). Final tier capped at Caution — cannot be lifted by any downgrade, example-payload opt-in, or allowlist.
Permissions & capabilities
No declared permissions — minimal attack surface.
network_incredential_accessnetwork_out Thanks — recorded.