The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze external context such as commit messages, pull request descriptions, and issue trackers to understand the intent of conflicting changes. These sources are user-controlled and may contain malicious instructions designed to subvert the agent's logic during the resolution process.
Ingestion points: Reading commit messages, PRs, and issues/tickets as described in SKILL.md.
Boundary markers: None specified to delimit external content or warn against embedded instructions.
Capability inventory: The skill performs git operations and executes automated project checks (SKILL.md).
Sanitization: No sanitization or validation of the external metadata is mentioned.
[COMMAND_EXECUTION]: The skill directs the agent to discover and run the project's automated checks, including typechecking, tests, and formatting tools.
Evidence: SKILL.md step 4 instructs the agent to "Discover the project's automated checks and run them."
These checks are defined within the repository's own configuration (e.g., package.json scripts, Makefile, or tox.ini) and can execute arbitrary code on the host system. Running these on an untrusted or compromised repository presents a risk of local code execution.

resolving-merge-conflicts