jobhunt-ops-broadcast-copy

Pass

Audited by Gen Agent Trust Hub on Mar 3, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect prompt injection surface.
  • Ingestion points: The skill is designed to ingest and parse updates from user messages and email screenshots as described in the 'Workflow' section of SKILL.md.
  • Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present when processing ingested content.
  • Capability inventory: The skill possesses file-write capabilities (updating STATUS.md, CHANGELOG.md, and JSON files in ~/.claude/) and command execution capabilities (running sync_db.ps1).
  • Sanitization: There is no evidence of sanitization or validation of the text extracted from screenshots before it is persisted into structured JSON or Markdown files.
  • [COMMAND_EXECUTION]: Execution of local scripts.
  • The skill executes jobhunt_ops/sync_db.ps1 during the persistence and status update phases to synchronize data between the workspace and the personal assistant database.
  • [DATA_EXPOSURE]: Access to application-specific hidden directories.
  • The skill reads and writes data to ~/.claude/personal_assistant/tasks.json and ~/.claude/personal_assistant/schedule.json. While these paths are hidden within the user's home directory, they appear to be intended storage for the skill's primary function as a personal assistant.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 3, 2026, 11:20 AM