Swarm Manager Review Agent

Compile plan, diff, and GCT context into verification evidence for backlog items.

54|7|Updated Sep 19, 2021
One-click install
npx skills add https://github.com/Vrooli/Vrooli --skill swarm-manager-review-agent
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Swarm Manager Review Agent
Source: https://github.com/Vrooli/Vrooli/tree/main/scenarios/prompt-manager/store/skills/packs/core/swarm-manager-review
Command: npx skills add https://github.com/Vrooli/Vrooli --skill swarm-manager-review-agent

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Human reviewers need trustworthy evidence to decide whether a completed backlog item should be archived or followed up, and this Skill compiles the plan, diff, and review context so nothing is guessed.

Core Features & Use Cases

  • Evidence Strategy Guidance: Walks through understanding the spec, diff summaries, changed paths, and affected scenarios before collecting evidence.
  • GCT-aware Validation: Evaluates available GCT review results to highlight green coverage, surface warnings, or prioritize red flags that touch the modified files.
  • Evidence Types & Suggestions: Recommends screenshots, API/CLI captures, config diffs, and manual assessments, then records suggestions for new automation that would let future rounds rely more on GCT.

Quick Start

Ask the Swarm Manager Review Agent to inspect the plan, diff, and GCT context and document the evidence so the reviewer can conclude the round.

Frequently Asked Questions about Swarm Manager Review Agent

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

FAQPage Schema
How do I collect verification evidence for completed backlog items?

To collect verification evidence for completed backlog items, compare planned requirements against actual diffs and contextual signals. This process uses plan.md, diff summaries, changed paths, and affected scenarios to document an evidence strategy for archive decisions.

What is proof-based backlog review and when do I need it?

Proof-based backlog review compiles plan, diff, and review context so human reviewers can trust the outcome. You need it when a finished execution requires an archive or follow-up decision based on trustworthy evidence rather than guesswork.

How do I validate backlog diffs using GCT review metrics?

To validate backlog diffs using GCT review metrics, evaluate available GCT results to highlight green coverage, surface warnings, and prioritize red flags touching modified files. This GCT-aware validation directs reviewer attention to critical changed paths.

What types of evidence do I need for a diff analysis review?

For a diff analysis review, recommended evidence types include screenshots, API captures, CLI captures, config diffs, and manual assessments. These captures document the evidence strategy alongside improvement suggestions for future automation.

Can I review backlog execution without GCT review metrics?

You can review backlog execution without GCT review metrics, as they are optional inputs. The review still proceeds by comparing plan.md requirements against diff summaries and changed paths, though it records suggestions for new automation to enable future GCT reliance.

How do I document an evidence strategy for backlog archive decisions?

To document an evidence strategy for backlog archive decisions, classify collected evidence by comparing planned requirements with actual diffs and affected scenarios. Honor attachment and schema requirements so reviewers can trust the outcome and conclude the round.