fleet-plan

Pass

Audited by Gen Agent Trust Hub on Apr 17, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill implements a dynamic discovery mechanism that reads FLEET-INDEX.md and individual SKILL.md files from the local directory to plan tasks. This content is then used to generate prompts for worker agents. This creates a surface for indirect prompt injection where malicious instructions in local repository files could influence the orchestrator's planning or the behavior of downstream workers.
  • Ingestion points: The skill uses Glob and Read tools to ingest FLEET-INDEX.md and SKILL.md files in Step 0 and Step 2.
  • Boundary markers: No explicit delimiters or instructions are provided to separate the ingested file content from the skill's own logic.
  • Capability inventory: The skill has Write, Edit, and restricted Bash (ls, mkdir) capabilities.
  • Sanitization: There is no evidence of validation or sanitization of the content retrieved from external files before it is processed.
  • [COMMAND_EXECUTION]: The skill instructs the user to execute shell scripts (launch.sh and status.sh) located in the skill's subdirectories to manage the lifecycle of the planned fleets.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 17, 2026, 11:09 AM