Kole By SyneHQ
v1.1.3Execute queries against your databases using SQL, PostgreSQL commands, or natural language through SyneHQ's Kole platform. Use this skill whenever the user m...
⭐ 1· 43·0 current·0 all-time
byHarsh Vardhan Goswami 🐳@theboringhumane
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
Name/description claim (execute SQL and NL queries via SyneHQ Kole) matches the declared requirements: the SKILL.md asks for SYNEHQ_API_KEY, optional SYNEHQ_CONNECTION_ID and an MCP server package (@synehq/kole-mcp). No unrelated env vars or binaries are requested.
Instruction Scope
The runtime instructions instruct the agent to use the MCP server to execute arbitrary SQL/DDL against connected databases and explicitly say to 'Always use this skill when user asks about their data, even if they don't explicitly say "database" or "SQL".' That broad trigger and the ability to run CREATE/ALTER/DROP are legitimate for a DB tool but increase the risk of accidental or unexpected destructive operations; the SKILL.md does warn to use least-privilege and test on non-production systems.
Install Mechanism
This is an instruction-only skill (no install spec in registry), but it directs users to install @synehq/kole-mcp from npm or GitHub. Installing a public npm package is a reasonable and expected install path, but you should review the package source (the SKILL.md itself recommends this) and prefer pinned versions and review package integrity before global installation.
Credentials
Only SYNEHQ_API_KEY (required) and SYNEHQ_CONNECTION_ID (optional) are requested. These are proportional to a service that must authenticate to query user databases. The SKILL.md warns that the API key grants access to all connected databases — treat it as highly sensitive and scope it to read-only when possible.
Persistence & Privilege
The skill does not request always:true or other elevated platform privileges. It instructs adding an MCP entry to .mcp.json (local config) which is normal for an MCP-backed skill. Autonomous model invocation remains enabled (platform default) but is not a new privilege requested by the skill itself.
Assessment
This skill appears to do what it claims: use SyneHQ's Kole to run SQL against your databases. Before installing or enabling it, do the following: (1) Review the @synehq/kole-mcp GitHub repository and the published npm package to ensure you trust the code and its maintainer. (2) Create and use least-privilege credentials (prefer read-only API keys or a separate audit/test connection) — the SYNEHQ_API_KEY can access all connected databases. (3) Do not store the API key in source control; use environment variables or a secrets manager. (4) Test the MCP server and the skill on non-production databases first to confirm behavior and avoid accidental DDL. (5) Consider pinning the MCP package version rather than installing @latest or installing globally; run it in an isolated environment if possible. (6) Because the SKILL.md instructs the agent to 'always' use the skill for any data-related question, be cautious: limit when the agent is permitted to run queries and monitor logs/audit trails so queries and DDL are visible. If you cannot review the MCP package or cannot restrict the API key, avoid installing the skill or use it only with sandbox/test databases.Like a lobster shell, security has layers — review code before you run it.
latestvk979t133eg8tqwvrhd5t6fzx8n84g6am
License
MIT-0
Free to use, modify, and redistribute. No attribution required.