Caveat verdict

wecom-agent-ops-center

🟢 Trusted

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

Provides observability and process guarding for AI agents in enterprise WeChat including secrets scanning and network whitelisting; auto-discovers local agents but message conversion through hermesai.ltd is disclosed and can be disabled.

⚠ 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

100

transparency

100

maintenance

Permission integrity

Installs packages at runtime — transitive dependencies are not auditable

package_install

Findings (19)

Pattern match critical

Possible hardcoded credential

SKILL.md · code · secret: "你的BotSecret

Pattern match critical

Pipe to python — executes piped content as Python code

TEST_GUIDE.md · code · | python3

Pattern match high

References sudo — requests elevated privileges

SKILL.md · code · sudo

Pattern match high

HTTP request to bare IP address — common in malicious payloads

ARCHITECTURE.md · code · http://10.0.1.10

Pattern match medium

Opens WebSocket connection

SKILL.md · frontmatter · websocket

Pattern match medium

References child_process — can spawn system processes

adapters/openclaw-adapter.js · prose · downgraded · child_process

Pattern match medium

Raw socket connection — low-level network access

agent-monitor.js · prose · downgraded · socket.connect(

Pattern match medium

Uses spawn() — can execute external programs

core/lifecycle-manager.js · prose · downgraded · spawn(

Pattern match medium

Uses exec() — may execute shell commands

core/secrets-scanner.js · prose · downgraded · exec(

Pattern match medium

Dynamic import() — loads module at runtime

dashboard-api.js · prose · downgraded · import('

Pattern match low

Python aiohttp session — async network access

agent_bridge.py · prose · downgraded · aiohttp.ClientSession

Pattern match low

Popular HTTP library — network access

agent-monitor.js · prose · downgraded · axios

Pattern match low

References agent memory files

agent-scanner.js · prose · downgraded · MEMORY.md

Pattern match low

Node http/https module — low-level network access

agent-scanner.js · prose · downgraded · require('https')

Pattern match low

Accesses sensitive environment variables

config.js · prose · downgraded · process.env.WECOM_BOT_SECRET

Pattern match low

References agent configuration files

config.py · prose · downgraded · AgentConfig

Pattern match low

Python os.getenv — reads environment variable

config.py · prose · downgraded · os.getenv(

Pattern match low

process.nextTick — defers execution to next tick

core/network-guard.js · prose · downgraded · process.nextTick(

Pattern match low

Changes file ownership

deploy-pairing-server.sh · 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_installnetwork_in

Is this flag fair?

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