skills/cskwork/handoff-skill/handoff/Gen Agent Trust Hub

handoff

Pass

Audited by Gen Agent Trust Hub on Nov 5, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXPOSURE_&_EXFILTRATION]: The skill implements proactive security measures by instructing the agent to 'Protect sensitive material' and 'Redact credentials, tokens, passwords, private personal data, and session cookies.' These are replaced with placeholders, preventing accidental exposure of secrets in the generated handoff files.
  • [COMMAND_EXECUTION]: While the skill instructions mention recording commands run or outputs observed, it does not mandate the execution of arbitrary or dangerous commands. It focuses on documenting existing activity to maintain context.
  • [PERSISTENCE]: The skill saves state to markdown files within the repository (docs/handoffs/) or the OS temporary directory. This is the primary and legitimate purpose of a handoff tool and does not constitute a malicious persistence mechanism.
  • [INDIRECT_PROMPT_INJECTION]: As the skill aggregates repository data and user requests into a single document, it possesses an inherent surface for indirect prompt injection. However, the instructions emphasize gathering 'factual state snapshots' and 'durable facts,' and the skill includes a 'Resume Check' to verify clarity and evidence-backed work, which serves as a manual/procedural mitigation.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 5, 2026, 04:14 AM