A specialized logging and recall system designed to record technical failures and provide actionable preventions for future tasks.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install failure-memory-log
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 failure-memory-log 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 Failure Memory?
Failure Memory is a robust framework for recording, analyzing, and learning from technical setbacks within an AI development environment. By systematically documenting every error, root cause, and resolution, this skill ensures that Openclaw Skills maintain a persistent memory of what went wrong and how it was fixed. It transforms passive error logs into an active knowledge base that agents can query to avoid falling into the same traps twice.
This skill is particularly valuable for complex development cycles where environment-specific bugs or dependency conflicts often recur. By integrating Failure Memory into your suite of Openclaw Skills, you create a self-improving loop where every resolved ticket or failed deployment strengthens the agent's ability to navigate future challenges successfully.
Failure Memory Use Cases
- Automatically logging shell command failures to document environment-specific fixes.
- Performing pre-task history checks to identify known pitfalls before starting a new migration or deployment.
- Generating structured failure reports to identify systemic weaknesses in a software architecture.
- Tracking recurring API integration issues that only appear under specific network or data conditions.
How Failure Memory Works
- Detection: The agent identifies a task failure or a non-zero exit code from a command execution.
- Logging: The skill records a structured entry including the category, context, exact error message, and the specific root cause discovered.
- Documentation: A resolution and a prevention strategy are added to the record to guide future interactions.
- Search: Before starting new operations, the agent searches the failures.md file using keywords related to the current task.
- Prevention: If a match is found, the agent alerts the user or adjusts its plan based on the documented prevention tip.
Failure Memory Setup
To initialize the tracking environment for these Openclaw Skills, execute the initialization script provided in the skill directory:
bash scripts/init.sh ./memory
This will create the necessary directory structure and the failures.md log file where all future error patterns will be stored.
Failure Memory Data Schema & Taxonomy
The skill organizes its historical data within a structured Markdown file to ensure compatibility with standard CLI tools and Openclaw Skills.
| Field | Description |
|---|---|
| Category | High-level classification (e.g., build, deploy, permissions, logic) |
| Context | Detailed description of what the agent was attempting to do |
| Error | The verbatim error message or symptom observed during the failure |
| Root Cause | The fundamental technical reason the error occurred |
| Resolution | The specific fix that resolved the issue |
| Prevention | Actionable advice to avoid the error in future sessions |
| Tags | Keywords used for optimized keyword and semantic search |
Failure Memory Advanced Features
- Pattern Recognition: Grouping failures by category to identify and suggest systemic fixes for repeat offenders.
- Pre-Task Recall: Automated grep-based search triggered at the start of any significant task to surface relevant history.
- Vector Search Support: Capability to integrate with memory_search for semantic retrieval of similar failure patterns even when keywords don't match exactly.
- Multi-Agent Learning: Sharing a centralized failures.md across different agents to build a collective organizational memory.
SKILL.md
Loading
Related Openclaw Skills
Clawd Presence
A terminal-based presence display that provides glancable status updates and visual monograms for AI agents in real-time.
ReefGram
A specialized social network transceiver allowing AI agents to share hardware telemetry and creative media with a neural network.
Cost Guardian
Cost Guardian is an advisory AI agent skill that prevents wasteful token spend by auditing cron jobs, polling intervals, and model selection.
Zoom Manager
A headless integration for managing Zoom meetings, recordings, and schedules directly through an AI agent or CLI.
Generect B2B Lead Generation & Email Discovery
A comprehensive API for real-time B2B lead discovery, company research, and AI-powered email finding and validation.
Generect API Lead Generation & Email Discovery
A powerful B2B intelligence tool for AI agents to find leads, companies, and verified professional emails in real-time.
Featured*
