Conductor, an extension for the Gemini CLI introduced in December, enables context-driven development by shifting project awareness into persistent, version-controlled markdown files for planning. On February 13, 2026, a new feature, "Automated Review," was released to enhance AI-assisted engineering safety and predictability. This feature extends Conductor's capabilities beyond planning and execution into validation, generating post-implementation reports on code quality and compliance.

The core methodology of the Automated Review introduces a rigorous "verify" step in the development lifecycle. After a coding agent completes its tasks, Conductor generates a comprehensive post-implementation report by integrating several analytical capabilities:

1. Code Review: Conductor performs deep static and logic analysis on newly generated files, acting as an automated peer reviewer. This goes beyond basic syntax checks to proactively identify complex issues such as race conditions in asynchronous blocks, potential null pointer dereferences, and logic errors that could lead to runtime exceptions. Technically, this involves advanced program analysis techniques like data flow analysis or control flow graph analysis to detect concurrency bugs, memory safety issues, and semantic deviations.

2. Plan Compliance: The system automatically cross-references the newly generated code against plan.md and spec.md files. It ensures that all phases outlined in the development roadmap have been addressed and that no critical requirements were omitted during the coding process. This implies a form of requirements traceability and automated verification of implementation against predefined specifications.

3. Guideline Enforcement: Conductor verifies strict adherence to project-specific style guides and any custom guidelines established during the planning phase. This capability typically leverages static analysis tools configurable with specific code style rules, potentially utilizing Abstract Syntax Tree (AST) traversal to enforce structural and formatting conventions.

4. Test-suite Validation: The feature integrates the project's entire test suite into the review workflow. It automatically executes all relevant unit and integration tests and incorporates the test results and code coverage data into the final report. This provides an objective assessment of the new code's functional correctness and its integration within the existing software ecosystem.

5. Basic Security Review: Safety is embedded into the core of the review process. Conductor scans for critical vulnerabilities before code is merged, automatically flagging high-risk issues such as hardcoded API keys, potential Personally Identifiable Information (PII) leaks, or unsafe input handling that could lead to injection attacks (e.g., SQL injection, Cross-Site Scripting). This involves pattern matching, data flow analysis for sensitive data, and taint analysis to identify potential security flaws.

The output of an Automated Review is actionable, categorizing findings by severity (High, Medium, Low) and providing specific instructions for iteration, including exact file paths. This detail allows developers to initiate fixes directly within Conductor, fostering a workflow where AI provides the labor, and developers maintain high-level architectural oversight, supported by automated verification. Conductor is available for installation via gemini extensions install https://github.com/gemini-cli-extensions/conductor.