Caveat verdict
sf-ai-agentscript
A comprehensive reference skill for Salesforce Agentforce Agent Script DSL development; no network access, credential handling, or suspicious behavior present.
⚠ 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.
Findings (5)
Uses eval() — can execute arbitrary code
references/grounding-multiagent.md · code · eval(
Prompt injection — tries to override agent instructions
references/fsm-architecture.md · prose · downgraded · Ignore all previous instructions
Pipe to bash — executes piped content as shell commands
references/preview-test-loop.md · prose · downgraded · | Bash
Pipe to python — executes piped content as Python code
references/preview-test-loop.md · prose · downgraded · | Python
Instructs covert action — may act without user awareness
references/asset-validation-profiles.md · prose · downgraded · silently
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class A). 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?
Thanks — recorded.