Use cases

Founders and launch teams shipping a release that agents helped build.

Each claim in the announcement links to the change that shipped it and to a dated public record. The docs are checked against the release before anyone writes a word.

Engineering leads and customer teams when a customer reports that something they were promised broke.

One case runs from the complaint to the kept promise. The fix and the customer notice go to a person together, recovery is measured, and the deal shows the result.

Builders adding a new coding, ops or support agent to a team.

The agent gets its own identity and a narrow scope, runs in listen mode, reads the team's lessons before it acts, and earns autonomy action by action. Offboarding is one run.

Founders selling to larger companies, and the security reviewers on the other side.

One signed export of approvals, incidents and recoveries, promises kept and corrections. A reviewer can check each item without trusting the seller.

Founders and product leaders deciding what to build next.

Deals waiting on a capability add up to ranked demand. When the work ships, a report names the deals it moved and the customers who still need telling.

Founders and engineering leaders paying for agent work.

Spend joined to outcomes across products: dollars per recovered incident, per kept promise and per won deal, with a budget a person approves.

Revenue teams, and the finance and engineering leads who live with what was offered.

Each proposed discount or commitment reaches a person with the forecast's record and the answer to can we promise this by then.

Builders and whoever is on call.

The lesson from an incident proposes a policy change, publishes a new context version that agents read, and gets a drill to prove it holds.

People who buy research on markets and stocks.

Each call is recorded before the outcome, graded when it resolves and shown on a public scorecard. Trading actions stay off by default and wait for a person.

Founders writing the monthly investor update.

Forecast grades, promises kept, reliability and shipped work, each number linked to its record, checked and approved before it is sent.

Builders choosing between agents or agent vendors.

Run two agents side by side in listen mode and compare their records per action: incidents, approvals needed and cost per outcome.

Anyone on the team, and the agents that work for them.

One question answered from every product's facts, each answer naming the product it came from and the date.

Customer teams and the customers they serve.

A shared page per customer with each promise, its state and what shipped, published only after a person approves it.

Security and platform leads when an agent key leaks or an agent starts doing harm.

One run revokes the agent version everywhere, lists what it touched, opens an incident and files a lesson.

Teams moving an agent to a new model or prompt.

The new agent version runs in a shadow test first and is compared with the old one on approvals, incidents and outcomes. A person approves the switch with the comparison attached.

Founders and sales engineers facing a security questionnaire.

Each answer is drafted from the trust packet with a link to its evidence. Questions the record cannot answer are flagged for a person, never guessed.

Compliance leads preparing for an audit of systems where AI agents act.

Record events mapped to the controls an auditor asks about, exported on a schedule, each event signed.

Account managers and customer success leads 60 days before a renewal.

A brief with promises kept and broken, incidents that touched the account and the forecast odds, approved before any of it reaches the customer.

Builders whose agents need outside docs, datasets or skills.

The agent buys a ContextFile artifact under a spending limit. The AI Source of Truth verdict shows before purchase, the artifact sits in quarantine until it passes a check, and larger purchases wait for a person.

Product teams that ship every week.

Notes built from shipped work, each line linked to its change, with the customers who were promised that work told in the same run.

Owners who want one page on what their agents did.

A weekly digest of actions, approvals, incidents, promises and spend, each line linked to its record.