Caveat verdict
opennexum
Receives external input AND uses eval
A multi-agent orchestration framework using OpenClaw ACP with webhook callbacks and a dispatch queue; all operations are local CLI commands and standard HTTP callbacks for the documented agent coordination purpose.
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.
Receives external input AND uses eval — the remote code-injection pattern (data-flow not verified)
LLM01 · LLM05 · ASI01 · ASI05
Findings (8)
HTTP request to bare IP address — common in malicious payloads
README.md · code · http://127.0.0.1
Dynamic Function constructor — equivalent to eval()
packages/spawn/src/status.ts · prose · downgraded · new Function(
Dynamic import() — loads module at runtime
packages/cli/src/__tests__/init.test.ts · prose · downgraded · import(`
References child_process — can spawn system processes
packages/cli/src/commands/callback.ts · prose · downgraded · child_process
Accesses OpenClaw config/secrets directly
README.md · prose · downgraded · ~/.openclaw/openclaw.json
References agent configuration files
SKILL.md · prose · downgraded · CLAUDE.md
Popular HTTP library — network access
packages/cli/src/commands/spawn.ts · prose · downgraded · got
Accesses sensitive environment variables
packages/cli/src/lib/webhook.ts · prose · downgraded · process.env.OPENCLAW_HOOKS_TOKEN
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.
network_indynamic_eval Thanks — recorded.