Conducts systematic research, data synthesis, and evidence-based analysis to support decision-making and documentation. Use when: analyzing topics, synthesizing research, gathering evidence, or investigating questions for documentation or strategy.
Manages the creation, versioning, deprecation, and maintenance of automation scripts across the workspace and templates. Use when: creating new scripts, updating script versions, deprecating scripts, or managing script dependencies in SCRIPTS.md.
Manages the creation, validation, and maintenance of skill files across the project. Use when: creating new skills, updating skill metadata, validating skill structure, or managing skill-agent mappings.
Guides the Engagement Leader through building a new AI agent team for any consulting domain. Covers requirements interview, web benchmarking, current team diagnosis, proposal generation, and user approval gate. Outputs an approved proposal JSON for scripts/team-builder.ts to execute. Triggered by: "build new agent…
Pulls the next waiting service ticket from the Phase A ticket queue and executes its referenced skill or script. Use when: processing the local service ticket queue, running "/ticket-run".
Scans, updates, and upgrades Bun dependencies and packages across the AI workspace (L0) or standalone project (L2/L3) while ensuring lockfile consistency, package.json version bumps, and security compliance. Use when: updating bun dependencies, upgrading packages to @latest, checking outdated packages in workspace or…
Upgrade an existing L2/L3 project to the current template version. Use when: upgrading a variant-based project, syncing template improvements, refreshing scripts/agents/skills/docs/commands.
Defines Zod runtime schema validation patterns, interface boundaries, and contract safety rules for multi-agent architecture, IPC channels, and API endpoints.
Runs before the agent uses a tool for Edit, Write, MultiEdit and Agent tool calls, executing gateguard-fact-force.ts and agent-model-gate.ts via bun with --mode (2 commands). From 5throck/ai-workspace-standards.