skills/iwanhe/pandawa/scrum-master/Gen Agent Trust Hub

scrum-master

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill parses user-provided JSON files to generate agile metrics and retrospective reports. This processes untrusted input into the agent's context. \n
  • Ingestion points: Sprint and retrospective data are ingested from JSON files specified by the user (e.g., sprint_data.json) via json.load() in all provided Python scripts. \n
  • Boundary markers: The skill does not explicitly define markers to isolate script output from the rest of the agent prompt. \n
  • Capability inventory: The skill facilitates local script execution using Python's standard libraries (statistics, argparse, math) to calculate velocity and health scores. No network or high-privilege file operations are performed. \n
  • Sanitization: Input text is processed using standard keyword matching and numerical calculations; no specific sanitization for prompt injection strings is performed on descriptive text fields. \n- [COMMAND_EXECUTION]: The skill instructs the agent to execute local Python scripts included in the package. These scripts utilize standard library modules and do not invoke dangerous shell commands or download external code. \n- [SAFE]: Extensive analysis across all categories (obfuscation, exfiltration, persistence, privilege escalation) found no evidence of malicious intent or hidden capabilities. The author's information and technical stack align with the professional project management use case.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 02:34 PM
Security Audit — agent-trust-hub — scrum-master