oma-coordination

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute the oma CLI tool and a local script located at .agents/skills/orchestrator/scripts/spawn-agent.sh to manage sub-agents. These commands are used to parallelize task execution and coordinate different project domains.\n- [PROMPT_INJECTION]: The workflow is exposed to indirect prompt injection due to its requirement to monitor status updates from other agents.\n
  • Ingestion points: The agent reads and polls progress-{agent}.md files to track the status of sub-tasks (SKILL.md, Step 3).\n
  • Boundary markers: The instructions do not define any delimiters or ignore-instructions for the content ingested from the progress files.\n
  • Capability inventory: The agent can spawn additional processes via the oma CLI and modify codebases using Serena MCP tools (SKILL.md, Logical Operations).\n
  • Sanitization: No validation or sanitization mechanisms are specified for the data read from external agent progress files.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 06:54 AM
Security Audit — agent-trust-hub — oma-coordination