skills.rest
teambrilliant
teambrilliantOFFICIAL
Explore all skills by this author →
WebsiteGitHub

retrospective

Analyze software development events to identify autonomy-limiting gaps and generate improvement tickets.

1|1|Updated Feb 25, 2026
Search Tags:#retrospective#post-mortem#process improvement#incident review#agent autonomy#learning loop
One-click install
npx skills add https://github.com/teambrilliant/tap-skills --skill retrospective-teambrilliant
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: retrospective
Source: https://github.com/teambrilliant/tap-skills/tree/main/skills/retrospective
Command: npx skills add https://github.com/teambrilliant/tap-skills --skill retrospective-teambrilliant

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill addresses the challenge of improving AI agent autonomy by systematically analyzing past events, identifying what hindered autonomous operation, and generating actionable improvements.

Core Features & Use Cases

  • Event-Driven Retrospectives: Initiates reflection based on specific triggers like feature shipping, incident resolution, or patterns of agent failures.
  • Autonomy-Focused Analysis: Analyzes events through the lens of context, harness, feedback, and scope gaps that limit agent independence.
  • Actionable Improvement Generation: Produces concrete learnings and creates improvement tickets to enhance agent capabilities.
  • Use Case: After a complex feature deployment where an agent struggled with autonomous task completion, run this retro to pinpoint the exact information gaps or tool limitations that caused the bottleneck and create tickets to fix them for future efficiency.

Quick Start

Run a retrospective for the recent incident that occurred yesterday.

Frequently Asked Questions about retrospective

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

FAQPage Schema
How do I run a retrospective to improve agent autonomy after an incident?▼

Run a retrospective by analyzing incident events to pinpoint context, harness, feedback, or scope gaps. This identifies what limited autonomous operation and generates concrete learnings with improvement tickets to enhance future agent performance.

What is an autonomy-focused post-mortem in software development?▼

An autonomy-focused post-mortem analyzes events like feature shipping or recurring agent failures to identify what hindered independent operation. It examines context, harness, feedback, and scope gaps to produce actionable improvements for AI agents.

When do I need to conduct a retrospective for AI agent failures?▼

Conduct a retrospective when patterns of recurring agent failures emerge, or after complex feature deployments where agents struggled with autonomous task completion. This process identifies exact information gaps or tool limitations causing bottlenecks.

Can I generate actionable improvement tickets from a post-incident review?▼

Yes, post-incident reviews generate actionable improvement tickets by analyzing specific events. This process produces concrete learnings and creates tickets to fix context, harness, feedback, or scope gaps that limited agent independence.

What is the best way to identify context gaps limiting agent independence?▼

The best way to identify context gaps is running just-in-time retrospectives. Analyzing events through the lens of context, harness, feedback, and scope gaps pinpoints exact limitations and generates actionable learnings to enhance autonomous performance.

Why does my AI agent struggle with autonomous task completion during feature deployment?▼

Agents struggle due to context, harness, feedback, or scope gaps. Running a retrospective after complex feature deployments pinpoints exact information gaps or tool limitations causing bottlenecks and creates tickets to resolve them.

Related Skills

Explore More Software Engineering Skills→
testdino-hqtestdino-hq
8

docs

Guide AI agents through TestDino MCP tools for test data access and analysis.

Official
Advanced
scanbottscanbott

agent-operating-pattern

Orchestrate agent lifecycle tasks with a fixed six-step execution cycle.

Community
Advanced
EmasoftEmasoft
1

amcos-emergency-handoff

Transfer critical work to an available agent during failures or blockers.

Community
Advanced
StudentCristianStudentCristian

nw-post-mortem-framework

Structure incident retrospectives with timelines, root-cause analysis, and action items.

Community
Advanced