# What is an agent audit trail?

Source: https://corespeed.io/glossary/agent-audit-trail

An agent audit trail records what each agent did, as whom, and how it ended. CoreSpeed keeps one per organization, in Activity: the action, the actor, the outcome and, for a failure, its error code. Outcomes are ok, error, held or denied. Prompts and sensitive request content are not copied into it. Every client writes to the same trail.

It rests on the documentation: [Activity & audit](https://corespeed.io/docs/activity)

At length, on the blog: https://corespeed.io/blog/what-is-an-audit-trail-for-ai-agents

## What people search for

- ai agent audit log
- ai agent logging
- agent audit trail
- what is an audit trail for ai agents

## Questions

### Does the trail keep your agent's prompts or reasoning?

No. Prompts are not audit payloads, and sensitive request content is redacted. A failure keeps only the text the agent saw, up to 512 characters.

### Why is a call missing from Activity?

It was refused before any tool ran, for an unknown tool or invalid arguments. Those calls leave no record.

### Who can read the trail?

Members see their own actions. Admins see the whole organization, and only admins see IP addresses and user agents.

### Is the cost shown in Activity the billed cost?

Yes. The dashboard joins each record to the billing ledger, which is the authority for money.

All terms: https://corespeed.io/glossary (Markdown: https://corespeed.io/glossary.md)

## Related pages

- [What is an AI agent identity? A principal of its own](https://corespeed.io/glossary/agent-identity): An agent identity is a principal of its own, with its own keys, an owner and reach limited to what is shared with it. Why a personal API key falls short.
- [What is human in the loop for AI agents?](https://corespeed.io/glossary/human-in-the-loop): Human in the loop means a person decides named agent actions before they run, under rules written in advance. How the card, the wait and the one run work.
- [Shared agent tools and controls for teams](https://corespeed.io/solutions/teams): Manage shared connections, approval policies, and tool usage for teams working across Claude Code, Codex, and other MCP clients.

## Sources

- Activity & audit: https://corespeed.io/docs/activity
- What is an audit trail for AI agents?: https://corespeed.io/blog/what-is-an-audit-trail-for-ai-agents

Reviewed 2026-11-14.

## Links

- Product overview: https://corespeed.io/ (Markdown: https://corespeed.io/index.md)
- Documentation: https://corespeed.io/docs
- Pricing: https://corespeed.io/pricing (Markdown: https://corespeed.io/pricing.md)
