/issues

Overview

/issues is a production command in the Operations category of the Prismatic Platform that provides intelligent issue tracking with automatic extraction capabilities and bidirectional GitLab synchronization. This command transforms the traditionally manual process of issue management into an automated, AI-assisted workflow that detects, categorizes, and tracks issues across the entire platform lifecycle.

Issue tracking in a platform with 99 umbrella applications, 400+ agents, and 2.8 million lines of code requires more than a simple ticketing system. The /issues command addresses this scale by integrating directly with the platform's code analysis tools, quality gates, and telemetry infrastructure to automatically extract issues from compilation warnings, test failures, quality gate violations, and runtime anomalies. These automatically extracted issues are synchronized with GitLab, creating a single source of truth for all platform work items.

This command operates under the L2+ authority level and is executed by the issue-tracking-specialist agent, which maintains deep understanding of the platform's issue taxonomy, priority classification system, and milestone structure. It is part of the platform's 216-command slash command registry, built on the AIAD (Autonomous Intelligence Agent Design) standard.

The command supports the full issue lifecycle: creation, assignment, status tracking, progress monitoring, resolution verification, and closure. It enforces the platform's mandatory session discipline protocol, which requires GitLab issues for all work items. The command also provides analytics capabilities, generating reports on issue velocity, resolution time distributions, and category trends.

Architecture

The issue tracking system integrates multiple data sources into a unified issue management pipeline.

+---------------------+     +-------------------+     +-------------------+
| Issue Extractors    | --> | Classification    | --> | GitLab Sync       |
| (Multi-Source)      |     | Engine            |     | (Bidirectional)   |
+---------------------+     +-------------------+     +-------------------+
  |  |  |  |                        |                         |
  |  |  |  +-- Runtime Telemetry    |                         v
  |  |  +----- Test Failures        v                 +-------------------+
  |  +-------- Quality Gates  +-------------------+   | Issue Store       |
  +----------- Manual Entry   | Priority Resolver |   | (ETS + GitLab)    |
                              | (P0-P4 + CVSS)   |   +-------------------+
                              +-------------------+           |
                                                              v
                                                    +-------------------+
                                                    | Analytics Engine  |
                                                    | (Velocity + Trend)|
                                                    +-------------------+

The Issue Extractors monitor multiple data sources for issue-worthy events. The Classification Engine categorizes issues by type (bug, feature, debt, security), domain (which app or subsystem), and impact. The Priority Resolver assigns priority based on severity, affected components, and strategic alignment with active milestones. GitLab Sync maintains bidirectional synchronization so that issues created or updated in either system are reflected in the other. The Analytics Engine computes velocity metrics, burndown charts, and trend analyses.

Usage

Issue Discovery and Listing

# List all open issues
/issues

# List issues filtered by priority
/issues --priority P0

# List issues for a specific app
/issues --app prismatic_perimeter

# List issues assigned to current session
/issues --mine

# Search issues by keyword
/issues --search "compilation warning"

Issue Creation

# Create a new issue interactively
/issues create

# Create with inline specification
/issues create --title "Fix memory leak in agent pool" --priority P1 --label bug

# Create from a quality gate violation
/issues create --from-quality-gate --gate dialyzer

# Create from test failure
/issues create --from-test-failure --test test/prismatic_web/live/dashboard_test.exs

# Batch create from all current violations
/issues create --from-violations --auto-classify

Issue Management

# Update issue status
/issues update 1234 --status in_progress

# Assign issue to an agent
/issues assign 1234 --agent elixir-core-specialist

# Add comment to issue
/issues comment 1234 "Root cause identified: race condition in GenServer init"

# Close issue with resolution
/issues close 1234 --resolution fixed --commit abc123

# Link related issues
/issues link 1234 1235 --type blocks

Analytics and Reporting

# Show issue velocity report
/issues analytics --velocity

# Show issue distribution by category
/issues analytics --distribution

# Show burndown chart for active milestone
/issues analytics --burndown --milestone "MVP Prismatic Perimeter"

# Export analytics report
/issues analytics --format markdown --output issue-report.md

Options & Parameters

Option	Type	Default	Description
(subcommand)	string	list	Action: `list`, `create`, `update`, `assign`, `comment`, `close`, `link`, `analytics`
`--priority`	string	all	Filter by priority: `P0`, `P1`, `P2`, `P3`, `P4`
`--app`	string	all	Filter by application name
`--label`	string	-	Filter or assign label: `bug`, `feature`, `debt`, `security`, `enhancement`
`--status`	string	open	Filter or set status: `open`, `in_progress`, `review`, `closed`
`--mine`	boolean	false	Show only issues assigned to current session
`--search`	string	-	Full-text search across issue titles and descriptions
`--milestone`	string	-	Filter by or assign to milestone
`--from-quality-gate`	boolean	false	Extract issues from quality gate violations
`--from-test-failure`	boolean	false	Extract issues from test failures
`--from-violations`	boolean	false	Extract issues from all current violations
`--auto-classify`	boolean	false	Automatically classify extracted issues
`--gitlab-sync`	boolean	true	Enable GitLab synchronization
`--format`	string	table	Output format: `table`, `json`, `markdown`, `csv`
`--output`	string	stdout	Output file for reports
`--verbose`	boolean	false	Show detailed issue information

Execution Flow

Source Identification: The command determines which data sources to consult based on the subcommand and options. For list operations, the issue store is queried. For create --from-* operations, the appropriate extractor is invoked.
Issue Extraction (for automated creation): Extractors scan the specified source (quality gates, test results, telemetry logs) and produce raw issue candidates with source evidence.
Classification: Each issue candidate is classified by the classification engine, which assigns type, domain, severity, and initial priority. Auto-classification uses heuristics based on the source type, affected files, and error category.
Deduplication: Issue candidates are checked against existing issues to prevent duplicate creation. Deduplication considers issue titles, affected files, and error signatures.
GitLab Synchronization: New issues are created in GitLab with appropriate labels, milestones, and assignments. Existing issues are updated if their status has changed locally. Remote changes are pulled and merged with local state.
Store Update: The local ETS-backed issue store is updated with the synchronized state.
Output Rendering: Results are formatted according to the --format option and displayed or written to the specified output.

Integration Points

Component	Integration Type	Description
Prismatic Agents	Execution	Executed by the `issue-tracking-specialist` agent
GitLab API	Synchronization	Bidirectional issue sync via GitLab REST API
Quality Gates	Extraction	Quality violations automatically generate issues
Telemetry	Extraction	Runtime anomalies can trigger issue creation
ETS	Storage	Local issue cache for fast queries
AIAD Registry	Discovery	Command registered via AIAD standard
Session Lifecycle	Enforcement	Mandatory issue tracking for all sessions
Milestone Tracking	Planning	Issues linked to strategic milestones

Best Practices

Use automatic extraction rather than manual issue creation whenever possible. Automatically extracted issues include structured evidence (stack traces, quality gate output, test results) that accelerates diagnosis and resolution.

Classify issues accurately at creation time. Incorrect classification leads to misallocation of effort and distorts velocity analytics. When using --auto-classify, review the assigned categories before confirming batch creation.

Link related issues to build a dependency graph of work items. The link types (blocks, relates-to, duplicates) help identify critical path dependencies and prevent parallel work on duplicate issues.

Use milestone assignment to align issues with strategic objectives. Unassigned issues tend to languish, while milestone-tracked issues receive appropriate prioritization and visibility.

Review analytics weekly. The velocity and distribution reports reveal patterns in issue creation (which subsystems generate the most issues) and resolution (which types take longest to close). These insights drive targeted quality improvement efforts.

Error Handling

Error Condition	Response	Recovery
GitLab API unreachable	Operates in local-only mode with sync retry	Check network, retry sync with `/issues --sync`
Duplicate issue detected	Shows existing issue with match confidence	Link as duplicate or update existing issue
Invalid priority assignment	Suggests valid priority with context	Use valid P0-P4 priority
Milestone not found	Lists available milestones	Select from available milestones
Rate limit exceeded	Queues remaining operations for batch retry	Wait for rate limit reset

Advanced Usage

Automated Issue Triage

# Run automated triage on all open issues
/issues triage --auto

# Triage with specific strategy
/issues triage --strategy severity-first --reassign

# Generate triage report without applying changes
/issues triage --dry-run --format markdown

Bulk Operations

# Bulk close resolved issues
/issues bulk-close --label resolved --milestone "MVP Prismatic Perimeter"

# Bulk reassign issues from one agent to another
/issues bulk-assign --from leon-cleaner --to elixir-core-specialist

# Bulk update labels
/issues bulk-label --add priority:P1 --filter "app:prismatic_perimeter status:open"

Doctrine Compliance

All commands operate under the NO MERCY, NO DOUBTS doctrine:

NO MERCY: Zero tolerance for incomplete execution or quality violations. Every quality gate violation, compilation warning, and test failure is eligible for automatic issue extraction. No quality problem goes untracked. The mandatory session discipline protocol requires GitLab issue tracking for all sessions -- sessions without issues are blocked.
NO DOUBTS: Full investigation before action, evidence-based results. Issues include structured evidence from their extraction source. Priority assignment is evidence-based, considering severity, affected components, and strategic alignment. Analytics reports are computed from verified data, not estimates.

/agents - List and manage agent ecosystem with status monitoring
/commit - Smart commit with quality gates and conventional format
/connect - MCP server connection management across 14+ servers
/code - Core coding implementation and feature development
/fix - Bug fix implementation with mandatory regression tests
/refactor - Safe refactoring with zero-regression guarantee

Connect & Contribute

Created by Tomáš Korcak (korczis) | Open Source under GHL

GitHub | GitLab | LinkedIn | Contact
Developer Portal | Architecture | Meet the Creator