Group Debug & Deploy Expert
ReviewAudited by ClawScan on May 14, 2026.
Overview
This DevOps skill has no executable code, but it gives broad operational authority while treating unprovided external rule files and a hard-coded project identity as authoritative.
Review this skill before installing. It is appropriate only if you want a high-authority DevOps assistant that can run commands and modify project files. Remove the bundled BeijiXing-AI project context, do not allow unreviewed repository rule files to become binding agent instructions, and require explicit approval before using credentials, SSH, or production deployment actions.
Findings (5)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
A repository or workspace file could influence how the agent behaves during deployment or debugging if the agent treats it as a higher-priority rule source.
The skill makes external, unprovided project rule files higher authority than the skill itself. In a debug/deploy context, treating workspace files as binding instructions can redirect the agent away from the user's intended task.
GOVERNANCE HIERARCHY: - TIER 0: [core_file_protection.md](../rules/core_file_protection.md) ← Supreme Law ... RULE: When conflicts arise, TIER 0 > TIER 1 > TIER 2
Do not let project-local rule files override user or platform instructions. Review or remove the external governance hierarchy before using the skill in arbitrary repositories.
If used carelessly, the skill could change files, run commands, deploy code, or interact with remote systems.
These permissions and tools are expected for a deployment/debugging skill, but they allow broad local changes, command execution, network operations, and SSH-based access.
permissions: - "file.read" - "file.write" - "terminal.execute" - "process.spawn" ... requires: bins: "git", "curl", "ssh", "python3", "java", "node", "npm"
Use it only in workspaces where command execution is intended, review commands before execution, and avoid running it against production systems without explicit approval.
Users may not know which accounts, tokens, SSH keys, or deployment credentials the skill expects to use.
The registry does not declare a primary credential, while capability signals indicate that sensitive credentials or OAuth tokens may be involved. This may be normal for deployment/SSH workflows, but the credential contract is not explicit.
Required env vars: none; Primary credential: none ... Capability signals: requires-oauth-token, requires-sensitive-credentials
Before use, define which credentials are in scope and require explicit confirmation before using SSH keys, OAuth tokens, cloud credentials, or production account access.
The agent may apply assumptions from the bundled BeijiXing-AI project to a different project, creating confusion or cross-project context leakage.
The skill is advertised as universal and multi-project, but it embeds a specific active project identity and project context. That persistent context can be reused or trusted in unrelated user projects.
instance_id: "PROJ-BJX-001" ... current_project: "BeijiXing-AI (北极星AI商机获客系统)" ... You are currently assigned to the BeijiXing-AI ... project instance
Remove or parameterize the bundled project-specific instance data before installing, and require per-project configuration rather than shipping an active project context.
Users may not be able to verify exactly which version or supporting files they are installing.
The provided manifest only contains LICENSE, SKILL.md, and VERSION, and the registry lists version 1.0.2 while SKILL.md/VERSION show 1.0.1. This is a provenance and packaging consistency issue, not evidence of malicious code.
Skill Framework: 1.0.1 ... Added marketplace metadata ... skill.json ... package.json ... Updated README-DEPLOY ... Regenerated SHA256 checksums
Publish a consistent package: align registry and file versions, include referenced files or remove the claims, and provide verifiable checksums if integrity is advertised.