dev-swarm-code-development

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [SAFE]: No security issues detected. The skill's behavior is consistent with a software development assistant performing local file and version control operations.
  • [COMMAND_EXECUTION]: The skill uses local git commands including git add, git reset, git commit, and git rev-parse to manage implementation and documentation changes within the repository. These are standard tools for the developer persona and are used appropriately within the workflow.
  • [PROMPT_INJECTION]: The skill contains a surface area for indirect prompt injection as it ingests content from external project files and has the capability to write to the filesystem and commit code. Evidence Chain: 1. Ingestion points: Reads from 10-sprints/, 04-prd/, 07-tech-specs/, and features/ directories. 2. Boundary markers: Absent. 3. Capability inventory: File-write access to {SRC}/ and features/ directories, and execution of local git commands. 4. Sanitization: Absent. Despite the surface, no malicious instructions were detected, and the skill includes internal guidelines to follow security standards like avoiding OWASP vulnerabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 08:28 PM