Merge Conflict Resolver is a skill that gives an AI agent a structured, 9-step framework for handling git merge conflicts rather than relying on unguided inference. The workflow begins with scope assessment: the agent counts conflicts and categorizes their types, such as add/add, modify/delete, or rename/delete scenarios, including binary file collisions. It then performs contextual analysis to understand the developer’s intent — whether the branch represents a feature, bug fix, or refactor — before selecting a resolution strategy. Strategies available are Accept Ours, Accept Theirs, or Manual Merge, chosen based on objective criteria derived from the conflict context. After applying a resolution, the skill runs language-specific verification steps tailored to Python (indentation consistency), JavaScript and TypeScript (duplicate import detection), Go (struct formatting), and SQL (foreign key consistency). A final validation pass greps the working tree for leftover conflict markers and can execute build and test commands to confirm the resolved files are functional. This approach addresses a documented failure mode in standard LLM-driven conflict resolution: truncated files, deleted logic, or broken syntax that results from asking a model to resolve conflicts without a disciplined process. The skill is language-agnostic at the workflow level while providing targeted checks for the languages listed above. It is a skill, not an MCP server, so it has no executable tools of its own and operates as a reasoning framework layered on top of an agent’s existing capabilities.

Use cases

Resolve git merge conflicts in an automated CI/CD pipeline without manual intervention

Handle rename/delete and add/add conflict scenarios that trip up naive LLM prompts

Apply language-specific post-resolution checks for Python, JS/TS, Go, or SQL codebases

Ensure no leftover conflict markers remain in files before committing a merge

Classify conflict types and select resolution strategy automatically based on branch intent

Validate that resolved code still builds and passes tests after conflict resolution

When to use it

When an AI agent is responsible for merging branches in an automated workflow

When the codebase includes Python, JavaScript, TypeScript, Go, or SQL files that need language-aware verification

When previous unstructured conflict resolution has produced broken or truncated files

When a clean, linear git history is required after automated merges

When not to use it

When the repository uses a version control system other than git

When the agent lacks the ability to read file contents and execute shell commands, since this skill depends on those underlying capabilities

When conflict resolution requires domain-specific business logic that cannot be inferred from the code context alone

When language-specific checks are needed for languages outside Python, JS/TS, Go, and SQL