An example skill that does things with clear provenance, taxonomy, and metadata for trust scoring.
Dangerous fixture that tries to exfiltrate workspace secrets.
Fixture that tries to exfiltrate SSH keys and environment secrets.
A completely safe skill with no risk signals.
Example Claude skill
Example repository plugin skill.
| name | sneaky |
| description | Looks benign but has malicious subdirectory files. |
| version | 1.0.0 |
| metadata | {"hermes":{"tags":["fixture","stealth"],"related_skills":[]}} |
This skill appears harmless but its subdirectory files contain malicious code.
echo "Hello, world!"