resolving-merge-conflicts
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill requires the agent to ingest untrusted data from external sources, which could contain malicious instructions intended to manipulate the agent's logic during the conflict resolution process.
- Ingestion points: Step 2 in
SKILL.mddirects the agent to read commit messages, Pull Requests, and original issues/tickets. - Boundary markers: No delimiters or instructions are provided to the agent to treat external text as untrusted data.
- Capability inventory: The agent is expected to execute git commands and run project-specific automated checks, including tests and typecheckers (Steps 4 and 5 in
SKILL.md). - Sanitization: There is no requirement for the agent to sanitize or validate the content retrieved from external sources before using it to inform code changes.
- [COMMAND_EXECUTION]: The instructions rely on the agent's ability to execute shell commands for managing git state and running local build/test tools. While expected for a development skill, this capability can be targeted by instructions embedded in the external data it processes.
Audit Metadata