Operation Log Query

View, filter, and analyze the complete operation history of your Content Nodes.

View audit logs

PuppyOne provides three ways to view audit logs: Dashboard, CLI, and API.

Dashboard

Open your Project and go to the Data page
Select the target file or folder
Click the Audit tab in the right panel
Review all recorded operations for that node

CLI

Use puppyone fs audit to inspect the audit log for any path:

# View the audit log for a file
puppyone fs audit /docs/readme.md
 
# View all operations under a folder
puppyone fs audit /docs
 
# Output JSON
puppyone fs audit /docs/readme.md --json

Example output:

PATH: /docs/readme.md
──────────────────────────────────────────────────────
TIME                  ACTOR              ACTION
──────────────────────────────────────────────────────
2025-03-09 14:30:00   [email protected]     update
2025-03-09 12:15:00   Research Agent     update
2025-03-08 09:00:00   [email protected]       create

API

Call the REST API to fetch audit logs for a specific node:

curl -X GET "https://api.puppyone.ai/api/v1/nodes/{node_id}/audit-logs" \
  -H "Authorization: Bearer YOUR_TOKEN"

Example response:

{
  "logs": [
    {
      "actor": "[email protected]",
      "action": "update",
      "node_path": "/docs/readme.md",
      "timestamp": "2025-03-09T14:30:00Z",
      "metadata": {
        "source": "dashboard",
        "version": 5
      }
    },
    {
      "actor": "Research Agent",
      "action": "update",
      "node_path": "/docs/readme.md",
      "timestamp": "2025-03-09T12:15:00Z",
      "metadata": {
        "source": "mcp",
        "version": 4
      }
    }
  ]
}

Log entry structure

Each audit log entry includes the following fields:

Field	Type	Description
`actor`	string	The actor identity - user email (for example `[email protected]`) or agent name (for example `Research Agent`)
`action`	string	Operation type; see the list below
`node_path`	string	Path of the Content Node being operated on
`timestamp`	string	Time of the operation (ISO 8601 format)
`metadata`	object	Extra information such as source and version number

Operation types

Operation Type	Description	Typical Scenario
`create`	Create a new node	Creating a file or folder; an agent generates new content
`update`	Modify node content	Editing a file; an agent updates data
`delete`	Delete a node	Removing a file or folder
`rollback`	Roll back to a historical version	Restore an earlier version
`sync`	Sync from an external data source	Connector sync from Notion, GitHub, etc.
`checkout`	Check out a node	Lock a node for editing
`commit`	Commit checked-out changes	Finish editing and release the lock

Filtering and search

The current puppyone fs audit CLI command is optimized for quick inspection and prints the full log directly. If you need more granular filtering by actor, time range, or action type, two approaches are recommended:

Use --json output and filter locally with jq or a script
Call the REST API directly and filter by fields in your own program

# Export JSON and filter it yourself
puppyone fs audit /docs/readme.md --json

Real scenarios

Scenario 1: Who modified this file last?

puppyone fs audit /specs/api-design.md

The most recent operation appears at the top of the log, so you can immediately see who changed it and when.

Scenario 2: What did an agent do in the last 24 hours?

puppyone fs audit / --json

After exporting the full log, you can filter for Content Writer entries from the last 24 hours in your own script or with jq to confirm it behaved as expected.

Scenario 3: Track all changes in a folder

puppyone fs audit /products

Inspect the complete operation history under /products. If you need a time range, export with --json and filter locally.

Scenario 4: Investigate bad data

When data looks incorrect, use the audit log to identify the change that introduced the issue:

# View the full operation history of the file
puppyone fs audit /data/pricing.json
 
# Pair it with version history to compare before and after
puppyone fs versions /data/pricing.json

Once you find the problematic change, you can restore the data with version rollback.

Next steps

Version Control - View and roll back file versions
Conflict Resolution - Handle concurrent edits from multiple actors
Auth for Agents - Control what agents are allowed to access

Overview Overview