A local, verifiable naming and storage layer that allows AI agents to stamp and log data retrieved from tools via the Model Context Protocol.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install can
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 can 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 CAN ASK MCP?
CAN ASK MCP integrates Clock Address Naming (CAN) with AI agents utilizing the Model Context Protocol (MCP). This implementation of Openclaw Skills provides a localized, verifiable naming and storage layer for data agents retrieve from various tools. By operating independently of the transport layer, it ensures that every piece of information an agent processes is stamped with a unique identity consisting of WHEN, WHERE, and WHAT.
By leveraging Openclaw Skills, agents can maintain a localized index of their history. This allows for the verification of data integrity without needing to trust the original source after the initial retrieval. It creates a robust foundation for agent memory, where data can be recalled and validated locally, ensuring the agent operates on precise, unmutated information.
CAN ASK MCP Use Cases
- Verifying the integrity of historical tool outputs before injecting them into an agent's context.
- Reducing API costs and latency by recalling previously tool-generated results from a local cache.
- Maintaining a permanent, searchable audit log of all interactions between an AI agent and its MCP tools.
- Enabling agents to function in offline environments by relying on verified local logs.
How CAN ASK MCP Works
- The AI agent initiates a request and receives a data payload from an MCP tool.
- The agent locally generates a SHA256 hash of the payload to create the WHERE identifier.
- A millisecond-precision timestamp is captured to define the WHEN component.
- A descriptive label or intent is assigned to the operation to serve as the WHAT component.
- Using Openclaw Skills, the agent appends this metadata (WHEN, WHERE, WHAT) to a local index.tsv file, creating a permanent and verifiable record.
CAN ASK MCP Setup
To implement this skill, ensure your environment has sha256sum and date utilities available. You can manually generate a CAN stamp for any tool output using the following commands:
WHEN=$(date +%s%3N)
WHERE=$(sha256sum <<< "tool_output_string" | awk '{print $1}')
WHAT="tool_name.result"
echo -e "$WHEN\t$WHERE\t$WHAT" >> ~/.can/index.tsv
CAN ASK MCP Data Schema & Taxonomy
CAN ASK MCP organizes data using a simple, high-performance tab-separated format within the local file system. This allows Openclaw Skills to quickly parse and verify entries.
| Column | Key | Description |
|---|---|---|
| 1 | WHEN | Unix Epoch timestamp in milliseconds. |
| 2 | WHERE | The SHA256 hash of the content, ensuring data integrity. |
| 3 | WHAT | A string describing the tool name, action, or metadata. |
All logs are typically stored in the ~/.can/index.tsv file, providing a centralized location for agent memory.
CAN ASK MCP Advanced Features
- Tamper Evidence: Automated validation checks that compare current data against the stored WHERE hash to detect mutation.
- Offline Recall: The ability to verify and use past results even when the MCP server or network is unavailable.
- Searchable Metadata: Advanced filtering of agent logs by intent or timestamp to optimize context retrieval.
- Zero-Integration Logging: Since the process is local, it requires no modifications to existing MCP servers or external tools.
SKILL.md
Loading
Related Openclaw Skills
Session Watchdog
Proactively monitors context levels to save critical session information before AI token compaction occurs.
session-wrap-up
An end-of-session automation tool that synchronizes code, extracts technical learnings, and evolves project rules using AI pattern detection.
Self-Hosted Whisper API
A local, privacy-focused speech-to-text solution that leverages OpenAI's Whisper model on Kubernetes for fast audio transcription and translation.
Commute Traffic Checker (TomTom)
A TomTom-powered AI skill that provides real-time traffic analysis, geocoding, and travel recommendations for any route.
URnetwork Proxy & Privacy Agent Skill
A decentralized privacy network skill that enables AI agents to access global content anonymously through secure HTTPS, SOCKS, and WireGuard proxies.
OpenClawLog
A comprehensive WordPress management skill for automating blog content, user registration, and media handling via XML-RPC and REST APIs.
METADATA
Featured*
