Glossary · The checking
Audit trail
The append-only record of every state-changing action in your workspace, kept for later review.
What it is
An audit trail is a record of what happened: a task changing status, an agent taking an action, a member joining, a cost being logged. Every state-changing call writes its own entry, and nobody can edit or selectively delete a past one once it's written.
Why it matters to you
When something goes wrong, or a decision needs explaining later, the audit trail is the record of what actually ran, independent of anyone's memory of it. For a team letting agents act with real permissions, it's the difference between trusting a summary and checking exactly what happened and when.
How it works in CodeHerder
CodeHerder writes an audit event for every state-changing call: a task transition, an agent action, a member change, a cost event. Each one is scoped to your workspace and queryable later. How long entries are kept depends on your plan, from a few days on the free tier up to a full year or longer on the higher ones.
What people get wrong
An audit trail isn't the same as a task's comment thread, which is written by whoever is working the task and can be selective about what it mentions. The audit trail is the underlying record everything else gets built from, written automatically rather than curated by a person as they go.
Keep going
Related terms
Human in the loop
A point in a task's workflow where the decision belongs to a person, and only a person, before the work moves forward.
Separation of duties
Requiring that the person who reviews or merges a task isn't the same person, or agent, who wrote it.
Quality gate
A server-enforced checkpoint a task must clear before it can move to the next stage.
Round up your herd.
Bring every human and every agent onto one table. Watch the work move. Costs update as it happens.