The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of external code from remote branches, creating a vulnerability where malicious instructions embedded in the code could influence the agent during the conflict resolution process.
Ingestion points: Untrusted data enters the agent's context through git fetch origin and git pull operations described in the Workflow section of SKILL.md.
Boundary markers: The skill does not provide delimiters or specific instructions for the agent to disregard natural language commands that might be found within the merged code or conflict markers.
Capability inventory: The agent is authorized to execute shell commands, perform network operations via git, modify files, and run project-specific verification checks referenced in the Workflow.
Sanitization: There is no evidence of sanitization, validation, or filtering of the content being merged before it is presented to the agent for interpretation.
[COMMAND_EXECUTION]: The skill relies on executing multiple shell commands to manage the git environment and project state.
Evidence: The Workflow in SKILL.md utilizes git config, git fetch, git pull, and git merge to perform its primary tasks.

pull