Caveat verdict
skill-security-reviewer
Encodes data AND uses eval
This is a malicious skill detection tool with anti-obfuscation capabilities; it performs read-only analysis with safe decoding and focuses on detecting how skills attack users rather than performing attacks itself.
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.
Encodes data AND uses eval — the obfuscated-execution pattern (atob + eval; data-flow not verified)
LLM05 · ASI05 · ASI10
Permission integrity
agent_memory
Findings (9)
Uses eval() — can execute arbitrary code
SKILL.md · prose · downgraded · eval(
Pipe to bash — executes piped content as shell commands
SKILL.md · prose · downgraded · |bash
Accesses AWS credentials file
SKILL.md · prose · downgraded · ~/.aws/credentials
Pipe to sh — executes piped content as shell commands
SKILL.md · prose · downgraded · |sh
Uses exec() — may execute shell commands
SKILL.md · prose · downgraded · exec(
Accesses .ssh directory
SKILL.md · prose · downgraded · .ssh/
Accesses cloud provider credentials
SKILL.md · prose · downgraded · ~/.aws
Accesses Kubernetes config (may contain cluster credentials)
SKILL.md · prose · downgraded · ~/.kube/config
References sudo — requests elevated privileges
SKILL.md · prose · downgraded · sudo
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class A/B/D). 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.
data_encodingdynamic_evalagent_memory Thanks — recorded.