subagent

Coordinate delegated subagents within an orchestrated workflow with ownership boundaries.

11|5|Updated Jan 9, 2026
One-click install
npx skills add https://github.com/rbergman/dark-matter-marketplace --skill subagent-rbergman
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagent
Source: https://github.com/rbergman/dark-matter-marketplace/tree/main/plugins/workflow/skills/subagent
Command: npx skills add https://github.com/rbergman/dark-matter-marketplace --skill subagent-rbergman

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill defines a safe protocol for delegated subagents to operate under an orchestrator, ensuring tasks are executed within clear ownership and with minimal return payload.

Core Features & Use Cases

  • Ownership boundaries: Subagents may touch only their own designated files and cannot modify shared state without approval.
  • Terse telemetry: Returns are compact and actionable; detailed traces go to history/.
  • Escalation & review: Blockers are escalated to the orchestrator for review and commit.

Quick Start

Activate a delegated subagent by passing its task to the orchestrator, which will manage session context and history logging.

Frequently Asked Questions about subagent

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I delegate tasks to subagents within an orchestrated workflow?

Coordinating delegated subagents within an orchestrated workflow requires scoping execution to designated files and enforcing strict ownership boundaries. This approach returns compact, actionable results while logging detailed traces to history, ensuring deterministic outcomes without unauthorized state modifications.

What is the best way to enforce file ownership boundaries for delegated subagents?

Enforcing file ownership boundaries for delegated subagents involves restricting their execution scope to explicitly designated files. This prevents unauthorized modifications to shared state and ensures that all changes remain isolated until reviewed and committed by the orchestrator.

How do escalation rules work when a subagent encounters a blocker?

Escalation rules for a blocked subagent require routing the issue directly to the orchestrator for review. The orchestrator then evaluates the blocker and manages the commit process, preventing the subagent from altering shared git state without explicit approval.

Can I keep git state under orchestrator control while running multiple subagents?

Keeping git state under orchestrator control while running multiple subagents is achieved by preventing subagents from directly modifying shared state. The orchestrator manages session context, reviews detailed history logs, and handles all commits to maintain deterministic workflow results.

Why does my subagent return verbose output instead of compact telemetry?

Subagents returning verbose output instead of compact telemetry lack a protocol for routing detailed traces to history. Implementing a terse telemetry approach ensures returns are compact and actionable, while comprehensive execution details are securely logged to history.