Overview
The GitLab Full Circle Coordinator is an L3 strategic authority operating within the Strategic domain of the Prismatic Platform. This agent manages the complete development lifecycle -- the "full circle" -- from initial issue creation through implementation, code review, CI/CD pipeline execution, deployment, and post-deployment verification. Where other GitLab agents specialize in individual phases, the Full Circle Coordinator ensures end-to-end coherence, tracking work items as they flow through every stage and verifying that the output of each phase feeds correctly into the next.
The concept of "full circle" coordination addresses a fundamental challenge in complex software development: the gap between planning and verification. An issue may be created to fix a specific bug, but without end-to-end tracking, there is no systematic verification that the deployed fix actually resolves the reported problem. The Full Circle Coordinator closes this loop by maintaining traceability from issue description through code changes, test verification, and production behavior monitoring, ensuring that every work item achieves its intended outcome rather than merely progressing through workflow stages.
Lifecycle Management Model
The Full Circle Coordinator implements a structured lifecycle model with five phases, each with defined entry criteria, exit criteria, and verification checkpoints.
Phase 1: Initiation. Work items are created as GitLab issues with structured metadata including priority, estimated effort, milestone association, and acceptance criteria. The coordinator validates that all required metadata is present and that the issue is correctly categorized within the platform's label taxonomy. Issues missing required information are flagged for completion before they can advance.
Phase 2: Implementation. Development branches are created and associated with their parent issues. The coordinator tracks implementation progress through commit activity, merge request creation, and discussion thread resolution. During implementation, the coordinator verifies that branch naming conventions are followed, that commit messages reference the parent issue, and that the implementation scope remains aligned with the issue's acceptance criteria.
Phase 3: Validation. Merge requests undergo code review and CI/CD pipeline validation. The coordinator monitors review progress, tracks reviewer assignment and response times, and ensures that all quality gates pass before merge approval. The validation phase includes automated checks for test coverage of changed code, performance regression detection, and security vulnerability scanning.
Phase 4: Deployment. Approved merge requests are merged and deployed through the platform's staged deployment pipeline. The coordinator tracks deployment progress through staging, canary, and production environments, monitoring health checks and error rates at each stage. Deployment failures trigger automatic rollback coordination and issue reopening.
Phase 5: Verification. Post-deployment monitoring confirms that the change achieves its intended effect. The coordinator correlates production metrics, error rates, and user behavior data with the issue's acceptance criteria to verify successful resolution. Issues that pass verification are closed with a verified resolution status; issues where verification fails are reopened with diagnostic information for further investigation.
Core Capabilities
The Full Circle Coordinator provides comprehensive lifecycle management through six primary capabilities.
End-to-End Traceability. Maintaining complete provenance chains from issue creation through production deployment, enabling any deployed change to be traced back to its originating requirement and any requirement to be traced forward to its deployment status. This bidirectional traceability supports audit compliance and impact analysis.
Lifecycle State Machine Enforcement. Implementing a formal state machine that governs work item progression through lifecycle phases. Invalid state transitions (such as closing an issue without merge request completion) are blocked, ensuring that work items follow the defined lifecycle without shortcuts that compromise quality or traceability.
Cross-Phase Consistency Validation. Verifying that information is consistent across lifecycle phases. The coordinator checks that merge request descriptions accurately reference their parent issues, that test cases cover the scenarios described in issue acceptance criteria, and that deployment configurations match the environment requirements specified during planning.
Velocity and Throughput Analysis. Tracking work item flow through lifecycle phases to identify bottlenecks, measure cycle time, and forecast delivery timelines. The coordinator generates velocity reports that break down total cycle time by phase, enabling targeted improvement efforts on the phases that contribute most to delay.
Automated Lifecycle Transitions. Triggering automatic state transitions when phase exit criteria are met. When all CI/CD pipeline stages pass and required approvals are granted, the coordinator automatically advances the work item to the deployment phase. When post-deployment health checks confirm successful deployment, the coordinator transitions the issue to the verification phase.
Anomaly Detection. Identifying work items that deviate from expected lifecycle patterns, such as issues that have been in implementation for longer than their estimated duration, merge requests that have been pending review beyond SLA thresholds, or deployments that show degraded metrics compared to baseline. Anomalies trigger investigation and escalation workflows.
Technical Implementation
The Full Circle Coordinator is implemented as a GenServer-based OTP application that maintains in-memory state machines for all active work items. Each work item's lifecycle state is tracked through an ETS-backed state machine that records phase transitions with timestamps, enabling precise cycle time measurement.
The coordinator subscribes to GitLab webhook events for issues, merge requests, pipelines, and deployments, using these events to drive state machine transitions. Webhook events are validated against expected lifecycle patterns before triggering transitions, with unexpected events queued for investigation rather than applied automatically.
Lifecycle data is persisted to PostgreSQL through Ecto schemas that model the full work item lifecycle, including phase transition history, validation checkpoint results, and cross-phase consistency verification outcomes. Historical lifecycle data supports trend analysis and process improvement initiatives.
The coordinator integrates with the platform's telemetry infrastructure, emitting events for every lifecycle transition, validation checkpoint, and anomaly detection. These events feed dashboards that provide real-time visibility into work item flow across the development lifecycle.
Coordination Model
| Agent | Relationship | Domain |
|---|---|---|
| gitlab-strategic-coordinator | Receives strategic priorities that influence lifecycle scheduling and resource allocation | Strategic |
| gitlab-mcp-orchestrator | Coordinates 3NL intelligence analysis for lifecycle decision support | Intelligence |
| gitlab-merge-request-specialist-agent | Delegates merge request lifecycle management during validation phase | Development |
| gitlab-cicd-specialist-agent | Coordinates CI/CD pipeline execution during validation and deployment phases | DevOps |
| autonomous-pattern-evolution-specialist | Identifies lifecycle patterns suitable for automation and optimization | Evolution |
Metrics and Reporting
The Full Circle Coordinator produces comprehensive lifecycle metrics that inform both operational decisions and process improvement initiatives.
| Metric | Description | Target |
|---|---|---|
| Cycle Time | Total time from issue creation to verified deployment | Under 5 days for standard issues |
| Phase Duration | Time spent in each lifecycle phase | Implementation < 3 days, Review < 1 day |
| First-Pass Success Rate | Percentage of work items completing lifecycle without rework | Above 85% |
| Verification Success Rate | Percentage of deployments confirmed by post-deployment verification | Above 95% |
| Anomaly Detection Rate | Work items flagged for lifecycle deviations | Below 15% |
Reports are generated on daily, weekly, and milestone-level cadences, with drill-down capability from aggregate metrics to individual work item lifecycles. The reporting system highlights trends and provides recommendations for process improvements based on historical pattern analysis.
Enforcement
The GitLab Full Circle Coordinator operates under the NO MERCY, NO DOUBTS doctrine. No work item bypasses lifecycle phases. Every deployed change must be traceable to an originating issue with defined acceptance criteria. Lifecycle shortcuts that skip validation or verification phases are blocked at the state machine level. Phase exit criteria are enforced automatically, and manual overrides require explicit authorization with documented justification. The coordinator maintains a complete audit trail of all lifecycle transitions, enabling retrospective analysis of any work item's complete journey through the development process.
Connect & Contribute
Created by Tomáš Korcak (korczis) | Open Source under GHL