launchdarkly-flag-discovery
Warn
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS] (MEDIUM): The marketplace.json file specifies a requirement for the @launchdarkly/mcp-server package, which is not hosted by a designated Trusted External Source.
- [PROMPT_INJECTION] (MEDIUM): The skill is vulnerable to indirect prompt injection (Category 8) because it processes untrusted metadata from the LaunchDarkly API. -- Ingestion points: External data is ingested via list-flags, get-flag, and get-flag-health tools. -- Boundary markers: There are no instructions to use delimiters or ignore embedded instructions within flag names or descriptions. -- Capability inventory: The skill generates actionable cleanup recommendations, meaning malicious metadata could influence an agent to recommend the deletion of critical infrastructure. -- Sanitization: No evidence of sanitization or validation of external content is present in the skill instructions.
Audit Metadata