subagents-workflow-augmentation

Automate parallel evidence collection and error localization for failed command-line tool runs.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/moraluco/Illusion-Cursor-dev-kit --skill subagents-workflow-augmentation
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagents-workflow-augmentation
Source: https://github.com/moraluco/Illusion-Cursor-dev-kit/tree/main/.cursor/skills/subagents-workflow-augmentation
Command: npx skills add https://github.com/moraluco/Illusion-Cursor-dev-kit --skill subagents-workflow-augmentation

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Provides a side-channel augmentation that parallelizes evidence gathering, error extraction, and pointer discovery for existing UE and Angelscript workflows so the authoritative command path remains unchanged and human review is minimized.

Core Features & Use Cases

  • Non-invasive augmentation: Runs shell and explore subagents in parallel while leaving original entrypoints (soft-ue-cli, Angelscript tests, retro workflows) untouched.
  • Standardized six-part output: Produces Command, Result, EvidencePaths, PrimaryError, Pointers, and NextStep to simplify triage and issue filing.
  • Recovery and boundary handling: Detects bridge/unreachable states, suggests recovery steps (editor launch, recheck), and surfaces offline-index artifacts and log paths for safe analysis.
  • Use cases: Automated post-test evidence capture for AngelScript TDD, automated source/document pointers for failed query commands, and templated retro entries for knowledge base write-back.

Quick Start

Ask the Kit to run subagents for the most recent failed soft-ue-cli or Angelscript test and return the six-part standardized augmentation output with pointers and next steps.

Frequently Asked Questions about subagents-workflow-augmentation

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

FAQPage Schema
How do I automate parallel evidence collection for failed command-line tool runs?

Automating parallel evidence collection for failed command-line tool runs is achieved by running shell and explore subagents that capture logs, extract primary errors, and surface document pointers alongside minimal recovery actions.

What is the standardized output format for automated error localization and triage?

The standardized output format for automated error localization is a six-part structure comprising Command, Result, EvidencePaths, PrimaryError, Pointers, and NextStep to simplify triage and issue filing.

Can I run subagents for post-test evidence capture without modifying my existing AngelScript TDD workflows?

Yes, you can run subagents for post-test evidence capture without modifying existing AngelScript TDD workflows because the augmentation operates as a non-invasive side-channel, leaving original entrypoints untouched.

How does evidence gathering handle unreachable states or offline-index artifacts during a retro postmortem?

Evidence gathering for retro postmortems handles unreachable states by detecting bridge conditions, suggesting recovery steps like editor launch, and surfacing offline-index artifacts and log paths for safe analysis.

What's the best way to parallelize diagnostics for failed soft-ue-cli online queries?

The best way to parallelize diagnostics for failed soft-ue-cli online queries is to run side-channel subagents that extract primary errors and source pointers while leaving the authoritative command path unchanged.