sweeping-defect-siblings

Pass

Audited by Gen Agent Trust Hub on Nov 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute several local scripts to reproduce and verify bugs, including python3 tests/sweep.py, probe.py, and check.py. These commands involve passing arguments (such as INPUT) derived from external reports.
  • [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to process untrusted data from external sources (bug reports and PR comments) to guide its reproduction and sweeping tasks, creating a risk surface for injection attacks.
  • Ingestion points: The agent ingests external bug reports and pull request comments, as described in the YAML description and Step 1 of the reproduction loop in SKILL.md.
  • Boundary markers: There are no boundary markers, delimiters, or explicit instructions provided to the agent to disregard malicious directives that might be embedded within the external report content.
  • Capability inventory: The skill provides the agent with the capability to execute shell commands and Python scripts (tests/sweep.py, probe.py, check.py), which could be abused if the agent is manipulated by injected instructions.
  • Sanitization: The workflow does not include any steps for sanitizing, validating, or escaping the content of the bug reports before using it to drive the agent's actions.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 14, 2026, 12:32 PM