Caveat verdict

sofagent

🟢 Trusted

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

Agent governance skill enforcing behavioral constraints including privacy protection and no unauthorized data transmission; its purpose is to constrain agent behavior.

⚠ 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.

security

transparency

maintenance

Findings (22)

Pattern match critical

Prompt injection — tries to override agent instructions

ignite.md · code · ignore previous instructions

Pattern match high

Pipe to sh — executes piped content as shell commands

audit/dist/log-reader.js · prose · downgraded · |sh

Pattern match high

Recursive delete from root or home — destructive command

audit/src/diff-parser.test.ts · prose · downgraded · rm -rf /

Pattern match high

Accesses sensitive system files

audit/src/diff-parser.test.ts · prose · downgraded · /etc/passwd

Pattern match high

Possible hardcoded credential

audit/src/rules/rule-a2.test.ts · prose · downgraded · apiKey = "${longKey}

Pattern match high

Uses eval() — can execute arbitrary code

audit/src/rules/skill-safety-rules.ts · prose · downgraded · eval(

Pattern match high

Dynamic Function constructor — equivalent to eval()

audit/src/rules/skill-safety-rules.ts · prose · downgraded · new Function(

Pattern match high

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

audit/src/skill-safety-check.test.ts · prose · downgraded · curl xxx | bash

Pattern match high

Pipe to bash — executes piped content as shell commands

audit/src/skill-safety-check.test.ts · prose · downgraded · | bash

Pattern match high

Accesses .ssh directory

ignite.md · code · .ssh/

Pattern match high

Uses spawn() — can execute external programs

loop-check.md · code · spawn(

Pattern match high

Pipe to python — executes piped content as Python code

scripts/benchmark.sh · prose · downgraded · | python3

Pattern match medium

References child_process — can spawn system processes

audit/dist/diff-parser.js · prose · downgraded · child_process

Pattern match medium

Uses exec() — may execute shell commands

audit/dist/log-reader.js · prose · downgraded · exec(

Pattern match medium

os.system/popen — direct OS command execution

audit/src/rules/skill-safety-rules.ts · prose · downgraded · os.system(

Pattern match medium

Base64 decode (atob) — may hide malicious payloads

audit/src/rules/skill-safety-rules.ts · prose · downgraded · atob(

Pattern match medium

References sudo — requests elevated privileges

scripts/install.sh · prose · downgraded · sudo

Pattern match medium

Accesses shell history/config

scripts/install.sh · prose · downgraded · ~/.zshrc

Pattern match low

Sets world-executable permissions

audit/src/rules/skill-safety-rules.ts · prose · downgraded · chmod 777

Pattern match low

Popular HTTP library — network access

audit/src/verify-evidence.test.ts · prose · downgraded · got

Pattern match low

References agent memory files

ignite.md · prose · downgraded · MEMORY.md

Pattern match low

References agent configuration files

ignite.md · prose · downgraded · CLAUDE.md

Why the tier is capped

Execution sink present in raw bytes (Hard Floor: class A/D/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