speckit-task-grounding

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect Prompt Injection surface. The skill ingests and processes untrusted documentation files (e.g., tasks.md, spec.md, plan.md) which could contain malicious instructions designed to influence the agent's validation report.
  • Ingestion points: The PowerShell script Validate-TaskGrounding.ps1 reads the full text of multiple markdown artifacts in the feature directory.
  • Boundary markers: The framework does not utilize specific delimiters or instructions to ignore or isolate embedded prompts within these artifacts.
  • Capability inventory: The skill utilizes file system access to read artifacts and generate reports using standard PowerShell commands.
  • Sanitization: Ingested content is processed using regular expressions and string matching, but no formal sanitization is applied before the content is interpolated into the final tasks.grounding.md report.
  • [COMMAND_EXECUTION]: Local script utilization. The skill includes and instructs the agent to execute specific PowerShell scripts (Validate-TaskGrounding.ps1 and Aggregate-TaskGrounding.ps1) to perform task extraction, evidence searching, and result aggregation. These scripts use native PowerShell functions such as Get-Content, Out-File, and Select-String to interact with the local filesystem.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 11:17 PM