Caveat verdict
safe-guardian
Writes files AND executes processes
A local blacklist/whitelist security layer that intercepts dangerous tool calls; installed locally via npm with no outbound exfiltration paths visible, and example code shows blocking, not enabling, dangerous commands.
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.
Writes files AND executes processes — may drop and execute malicious scripts
LLM05 · LLM06 · ASI05
Installs packages AND executes processes — opaque dependency chain with execution
LLM03 · ASI04
Accesses credentials AND writes files — may persist stolen credentials locally
LLM02 · LLM06 · ASI03
Both reads and writes files — verify scope is limited to intended directories
LLM06 · ASI02
Permission integrity
credential_access
package_install
Findings (7)
Recursive delete from root or home — destructive command
SKILL.md · code · rm -rf /
Possible hardcoded credential
SKILL.md · code · apiKey: 'your-api-key
Accesses sensitive system files
SKILL.md · code · /etc/shadow
Uses exec() — may execute shell commands
SKILL.md · code · exec(
References child_process — can spawn system processes
SKILL.md · code · child_process
Sets world-executable permissions
config/blacklist.json · prose · downgraded · chmod 777
Changes file ownership
config/blacklist.json · prose · downgraded · chown
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class 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.
package_installcredential_accessfile_readfile_writeprocess_exec Thanks — recorded.