Caveat verdict
glm-4.6v-vision-connector
visual-understanding
Reads local files AND makes external network calls
GLM-4V vision model integration that encodes images to base64 and sends them to Zhipu AI API; explicitly documents security requirements (no hardcoded keys, PII scrubbing) and behavior matches stated image analysis purpose.
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.
Reads local files AND makes external network calls — the capabilities for data exfiltration co-occur (data-flow not verified)
LLM02 · LLM06 · ASI03
Reads files, encodes data, AND makes external network calls — the obfuscated-exfiltration pattern (data-flow not verified)
LLM02 · ASI03
Accesses credentials AND makes external network calls — potential credential theft
LLM02 · ASI03
Accesses credentials AND encodes data — may obfuscate stolen credentials
LLM02 · ASI03 · ASI04
Permission integrity
network_out
file_read
credential_access
package_install
Findings (1)
Python os.environ.get — reads environment variable
SKILL.md · code · os.environ.get(
Permissions & capabilities
No declared permissions — minimal attack surface.
package_installfile_readcredential_accessdata_encodingnetwork_out Thanks — recorded.