aspire-project-v2-migration
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- Untrusted Content Processing (Indirect Prompt Injection): The skill analyzes local source code, project metadata, and configuration files to build its migration assessment. This ingestion of untrusted data from the workspace creates a potential surface for indirect prompt injection where malicious instructions could be embedded in the code being analyzed.\n
- Ingestion points: Workflow Step 1 in
SKILL.mddescribes the inspection of the application workspace.\n - Boundary markers: The skill establishes a hard gate requiring explicit user approval of a per-resource plan before any edits occur.\n
- Capability inventory: The skill can modify project files and execute CLI commands for orchestration.\n
- Sanitization: The skill relies on presenting a clear mapping table to the user for validation before any action is taken.\n- Project-Specific Configuration Deferral: The skill includes instructions to check for a project-local configuration file in the
.agents/skills/directory. While this supports extensibility, it identifies a mechanism where local repository files can influence agent behavior. The instructions mitigate this risk by requiring the agent to warn the user and retain its internal safety gates if a local override is used.\n- Developer Tool Orchestration: The skill interacts with standard command-line utilities for documentation lookup and application orchestration. These operations are used to verify the integrity of the migration and represent standard functionality for developer-focused utilities.
Audit Metadata