bulk-password-update

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the processing of untrusted data from the input CSV.
  • Ingestion points: The skill reads the 'label' and 'uid' fields from a user-provided CSV file and presents them in the user interface (SKILL.md).
  • Boundary markers: There are no explicit delimiters or instructions provided to the agent to distinguish the data content from instructional commands.
  • Capability inventory: The skill has the capability to execute shell commands (nfc-list, nfc-mfultralight, nfc-mfclassic) and perform file system writes (.progress.json, write-log.jsonl) (SKILL.md).
  • Sanitization: There is no evidence of sanitization or validation performed on the CSV fields before they are output or used in the workflow.
  • [COMMAND_EXECUTION]: The skill executes command-line utilities to interact with hardware.
  • Evidence: The skill relies on nfc-list, nfc-mfultralight, and nfc-mfclassic to detect tags, authenticate, and write new passwords (SKILL.md). These capabilities are appropriate for the skill's purpose but represent a surface that could be targeted by injection vectors.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 10:52 AM
Security Audit — agent-trust-hub — bulk-password-update