working-memory-boost

Pass

Audited by Gen Agent Trust Hub on Apr 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the GitHub CLI (gh) to perform remote state synchronization. It executes commands to view and update GitHub Issues to serve as a persistent remote storage layer for local task data.\n- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests and processes untrusted data from remote GitHub Issues to reconstruct the local environment state.\n
  • Ingestion points: Content from GitHub Issue bodies and comments is retrieved and parsed in SKILL.md (specifically during the resume flow).\n
  • Boundary markers: No boundary markers or delimiters are used to isolate the external issue content from the agent's instructional context.\n
  • Capability inventory: The skill has permissions to write to the file system, manage directories, and execute Git commands based on the parsed state.\n
  • Sanitization: There is no evidence of validation or sanitization of the remote content before it is used to rebuild local files and state.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 10, 2026, 09:26 PM