Caveat verdict

swift-mlx-lm

mlx-swift-lm

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

Receives external input AND uses eval

Reference documentation for the MLX Swift LM library covering local inference, LoRA fine-tuning, and embeddings on Apple Silicon; no network calls or exfiltration patterns present.

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
80
transparency
70
maintenance

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.

Capability combination critical

Receives external input AND uses eval — the remote code-injection pattern (data-flow not verified)

LLM01 · LLM05 · ASI01 · ASI05

Capability combination critical

Encodes data AND uses eval — the obfuscated-execution pattern (atob + eval; data-flow not verified)

LLM05 · ASI05 · ASI10

Findings (3)

Pattern match critical

Uses eval() — can execute arbitrary code

SKILL.md · code · eval(

Pattern match critical

Possible hardcoded credential

references/model-container.md · code · Token: "your_token

Pattern match medium

Popular HTTP library — network access

references/tool-calling.md · code · got

Permissions & capabilities

No declared permissions — minimal attack surface.

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