agent-coordination
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes output from subagents via system events and local coordination files, which presents an attack surface where a compromised subagent could attempt to inject instructions. * Ingestion points: Agent tool results and files in the
.coordination/directory (SKILL.md, references/fallback-coordination.md). * Boundary markers: While the skill suggests prompt structures for tasks, it does not specify methods for sanitizing or escaping the data returned by subagents (references/claude-code-coordination.md). * Capability inventory: The skill utilizes the Agent tool to spawn subagents and performs filesystem operations for coordination (SKILL.md, references/fallback-coordination.md). * Sanitization: No explicit validation or sanitization rules for subagent-provided data are mentioned. - [COMMAND_EXECUTION]: The README provides a setup command using
npxto install additional agent skills (npx skills add jwilger/agent-skills), which executes code from a remote source (SKILL.md). - [EXTERNAL_DOWNLOADS]: The skill references and facilitates the download of related modules from the author's GitHub repository at
jwilger/agent-skills(SKILL.md).
Audit Metadata