Security audit
Security checks for vulnerabilities and agentic risk
This skill provides a bounded search helper with disclosed Tavily network search and local memory/skills search that are both gated by explicit flags.
Install only if you are comfortable with a search tool that can send explicitly approved queries to Tavily and can read workspace memory/ and skills/ when explicitly enabled. Avoid using --allow-network for confidential or proprietary queries, and treat returned web results as evidence to verify rather than final truth.
The skill description overstates its capabilities by implying stronger provenance guarantees and broader source support than are actually implemented. Security-sensitive workflows may trust the output as evidence-grade or assume certain search backends are active, leading to incorrect decisions, misplaced trust, or failure to apply additional verification where needed.
The skill advertises network and environment-dependent behavior but does not declare an explicit tool scope such as permissions or allowed-tools. That creates a policy and review gap: a caller may invoke the skill without clear visibility into its access requirements, increasing the chance of unintended network use or secret exposure through environment-backed integrations.
The hardcoded Tavily endpoint indicates intentional outbound data transfer to a third-party service. In this skill's context, that is expected functionality, but it still creates a real privacy and exfiltration risk because search terms may contain proprietary or sensitive information, and the only safeguard is a narrow keyword check plus a runtime flag.
if (!ALLOW_NETWORK) return { items: [], warning: 'Network search disabled; pass --allow-network after confirming the query may be sent to Tavily' };
if (/\b(password|passwd|secret|token|api[_ -]?key|private key|credential)\b/i.test(query)) return { items: [], warning: 'Network search blocked for a potentially sensitive query' };
if (!process.env.TAVILY_API_KEY) return { items: [], warning: 'TAVILY_API_KEY is not set' };
const response = await fetch('https://api.tavily.com/search', {
method: 'POST', headers: {'content-type': 'application/json'},
body: JSON.stringify({ api_key: process.env.TAVILY_API_KEY, query, max_results: limit, search_depth: 'basic', include_answer: false }),
signal: AbortSignal.timeout(20_000)
The hardcoded Tavily endpoint indicates intentional outbound data transfer to a third-party service. In this skill's context, that is expected functionality, but it still creates a real privacy and exfiltration risk because search terms may contain proprietary or sensitive information, and the only safeguard is a narrow keyword check plus a runtime flag.
if (!ALLOW_NETWORK) return { items: [], warning: 'Network search disabled; pass --allow-network after confirming the query may be sent to Tavily' };
if (/\b(password|passwd|secret|token|api[_ -]?key|private key|credential)\b/i.test(query)) return { items: [], warning: 'Network search blocked for a potentially sensitive query' };
if (!process.env.TAVILY_API_KEY) return { items: [], warning: 'TAVILY_API_KEY is not set' };
const response = await fetch('https://api.tavily.com/search', {
method: 'POST', headers: {'content-type': 'application/json'},
body: JSON.stringify({ api_key: process.env.TAVILY_API_KEY, query, max_results: limit, search_depth: 'basic', include_answer: false }),
signal: AbortSignal.timeout(20_000)
Detected: suspicious.env_credential_access