A structured framework for AI agents to evaluate their performance, extract reusable patterns, and improve future task execution.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install task-retrospective
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install task-retrospective using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Task Retrospective?
The Task Retrospective skill provides a comprehensive methodology for AI agents to engage in post-action analysis. By reflecting on completed workflows, agents can identify what worked, what failed, and how to optimize their decision-making processes. Using this Openclaw Skills framework ensures that every completed task serves as a learning opportunity, transforming one-off successes into repeatable high-performance patterns.
It bridges the gap between simple execution and continuous improvement, allowing for a more sophisticated and self-aware agent lifecycle. By formalizing the evaluation of efficiency, accuracy, and strategy, this skill turns raw task data into actionable intelligence for future sessions.
Task Retrospective Use Cases
- Analyzing the outcome of complex multi-step coding or debugging sessions to find faster paths.
- Evaluating the efficiency of research and planning phases after completing a long-term project.
- Building a personal pattern library based on recurring successes and failures across different domains.
- Reviewing multi-agent handoffs to identify bottlenecks and communication gaps in collaborative workflows.
How Task Retrospective Works
- Reconstruct the task timeline by reviewing all steps taken, tools utilized, and major decisions made during execution.
- Evaluate the performance across specific dimensions including efficiency, accuracy, and the quality of the chosen approach.
- Extract valuable insights such as reusable patterns, identified anti-patterns, and key decision heuristics.
- Record a structured retrospective report in Markdown for future reference and long-term agent improvement.
Task Retrospective Setup
To implement this within your environment, ensure your agent is prompted to follow the structured evaluation format. You can initiate a review using standard Openclaw Skills triggers:
# Basic command to start a review
Run a retrospective on the task I just completed.
For more detailed analysis, provide specific context regarding the outcome and unexpected findings:
# Detailed command for specific tasks
Retrospective: [task description]. Outcome: [success/failure]. Time spent: [duration].
Task Retrospective Data Schema & Taxonomy
The skill generates a structured retrospective report that organizes metadata and insights into the following taxonomy:
| Section | Description |
|---|---|
| Summary | 1-2 sentence overview of the task and its final outcome. |
| Timeline | A table tracking duration and verdict for Research, Planning, Execution, and Validation. |
| What Worked | A list of successful patterns that should be repeated in similar contexts. |
| What Didn't Work | Analysis of anti-patterns identified and the better alternatives to use next time. |
| Reusable Patterns | Named strategies with detailed descriptions on when and how to apply them. |
| Key Decisions | Documentation of choice points, the decision made, and the resulting impact. |
| Improvement Actions | A checklist of specific, actionable steps to enhance future agent performance. |
Task Retrospective Advanced Features
- Approach Comparison: Benchmarking current actions against ideal theoretical workflows to identify optimization gaps.
- Pattern Library Integration: Aggregating insights across multiple sessions to identify long-term trends and recurring success factors.
- Multi-Agent Support: Specialized retrospectives for team-based workflows focusing on handoff points and collective bottlenecks.
- Behavioral Analytics: Tracking which improvement actions are consistently followed to measure agent growth over time.
SKILL.md
Loading
Related Openclaw Skills
Vulnerability Prioritizer
A high-performance risk prioritization engine that filters scanner noise using real-world exploit intelligence and asset context.
Monorepo Analyzer
A comprehensive auditing tool for identifying workspace structures, internal dependencies, and architectural issues in monorepos.
Backpressure Analyzer
A diagnostic and optimization tool for detecting processing bottlenecks and implementing flow control in data streaming pipelines.
GitHub Actions Linter
A specialized linter for GitHub Actions YAML files to detect security risks, deprecated actions, and syntax errors.
KAMIS CLI: Korean Agricultural Price Data
A specialized CLI tool for fetching daily wholesale and retail price data for Korean agricultural, livestock, and aquatic products via the official KAMIS OpenAPI.
Prettier Config Validator
A robust diagnostic utility for validating and linting Prettier configuration files across multiple formats to ensure project consistency.
Featured*
