launchdarkly-flag-targeting

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE] (SAFE): Analysis of the skill's instructions and references found no evidence of malicious intent, data exfiltration, or unauthorized command execution.
  • [Indirect Prompt Injection] (LOW): The skill exhibits a surface for indirect prompt injection as it ingests data from external sources that may be user-controlled.
  • Ingestion points: Data retrieved from the LaunchDarkly API via the get-flag tool (referenced in SKILL.md and safety-checklist.md).
  • Boundary markers: No explicit delimiters or boundary markers for external data are defined within the skill instructions.
  • Capability inventory: Significant modification capabilities including toggle-flag, update-rollout, update-targeting-rules, and copy-flag-config (defined in SKILL.md).
  • Sanitization: No specific sanitization or validation of the ingested flag metadata (such as descriptions or comments) is described in the skill logic.
  • Assessment: The risk is mitigated by the skill's strong focus on human-in-the-loop safety checklists, environment verification, and support for LaunchDarkly's native approval workflows.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:25 PM