ClawHub
Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Ops Skills Pack

v1.0.0

Provides production-ready autonomous agent operations with cost optimization, task autonomy, persistent memory, security, and scheduled execution workflows.

⭐ 0· 269·1 current·1 all-time
byEric JIE@erich1566
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Suspicious
medium confidence
ℹ
Purpose & Capability
The skill pack's name, description, and many included .md files align with an operations/ops-purpose (model routing, cron orchestration, file-based memory, logging). It does not request any environment variables or external credentials, which is consistent for an instruction-only doc package — however the docs recommend configuring commercial model identifiers (e.g., anthropic/claude-sonnet-4-6, openrouter/...) but do not mention obtaining or supplying API keys. Minor provenance inconsistencies appear (multiple GitHub URLs and 'yourusername' placeholders).
ℹ
Instruction Scope
Runtime instructions are purely filesystem and OpenClaw CLI actions: copying .md files to ~/.openclaw/workspace, creating USER.md/MEMORY.md/Todo.md/SOUL.md, editing ~/.openclaw/config.json, and adding cron jobs via openclaw cron. They do not prompt reading unrelated system files or sending data to external endpoints. However they instruct creating persistent workspace files (memory, user profile, lessons) which may contain sensitive information if the user puts secrets there; and they recommend using OpenClaw commands like 'openclaw security audit --fix' and 'openclaw doctor --deep --fix --yes' which can automatically modify system/agent state — the docs provide no explicit safety checks for those automatic fixes.
✓
Install Mechanism
This is an instruction-only pack with no installer or archive downloads; nothing is written to disk by an installer beyond what the user explicitly copies. That keeps install risk low. Note: package.json is present but there is no install spec in the registry; running npm scripts from this repo is not required by the SKILL.md but could be misleading.
ℹ
Credentials
The package declares no required environment variables or credentials (proportionate for a docs-only skill). That said, the configuration examples reference third-party model providers (Anthropic/openrouter) — using those models in practice will require provider API keys and billing in your OpenClaw configuration, yet the docs do not explain or request these credentials. Also, user/workspace files (USER.md, MEMORY.md) may be used to store identity or state; users should not place secrets there.
!
Persistence & Privilege
The pack intentionally enables persistent, autonomous behaviors: file-based memory (MEMORY.md, STATE.md, etc.) and cron-orchestration that schedules regular autonomous runs. By itself this is consistent with the stated goal (overnight autonomous work), but persistent local files plus scheduled autonomous execution increases the blast radius if misconfigured — e.g., sensitive data stored in memory files, cron messages that cause broad changes, or unreviewed automatic 'fix' commands. The skill is not always: true, but autonomous invocation + scheduled jobs can run without per-run user confirmation if OpenClaw is configured that way.
What to consider before installing
What to check before installing: - Review the .md files locally before copying them into ~/.openclaw/workspace. The package is documentation-heavy: copying without review gives your agent new behavior and persistent files immediately. - Do not store secrets, API keys, or passwords in USER.md, MEMORY.md, STATE.md or similar workspace files. Treat these as plaintext agent-visible memory. - The docs reference commercial models (Anthropic, OpenRouter). Using those models will require provider API keys and may incur billing; the package does not request or configure those keys for you. - The package suggests commands like 'openclaw doctor --deep --fix --yes' and 'openclaw security audit --fix' — automatic '--fix' actions can change agent/system state. Run audits in dry-run or review mode first. - The pack enables scheduled autonomous runs (cron). If you enable cron jobs, verify each cron message and the agent's scope-control/safety policies so background runs cannot make unwanted changes. - package.json references install/verify scripts but there are no scripts/JS files included. Avoid running npm install or the package scripts from this repo unless you inspect their contents; they may fail or be misleading. - Verify the repository/source and author: the repo URLs and placeholders are inconsistent in places. If provenance matters for your deployment, confirm the upstream source and review change history. - If you want to test safely: install to a sandbox OpenClaw workspace first (not your production workspace), disable automatic fixes, and monitor the first few overnight runs before trusting the pack in production.

Like a lobster shell, security has layers — review code before you run it.

Operationsvk97f2qnmhmve0308ebdg20cj2x82m53clatestvk97f2qnmhmve0308ebdg20cj2x82m53c

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Comments