Before: A draft finding extrapolates beyond the tested sample.
Evidence limits the conclusion
Sample facts and procedure results stay visible, so unsupported population-wide language is removed or qualified.
HINT for Auditors
Your assistant drafts findings fast. HINT keeps conclusions inside the evidence.
HINT keeps engagement scope, criteria, procedures and evidence in plain-text notes beside working papers. Your assistant reads only what applies to this engagement before drafting a finding.
Where it helps
Three situations your team will recognize
Before: Severity is chosen because the wording sounds serious.
The declared scale decides severity
Impact and likelihood must map to the engagement's approved rating criteria.
Before: A useful observation drifts outside engagement scope.
Scope remains a wall
The assistant separates an out-of-scope issue for follow-up instead of presenting it as an audited conclusion.
See it
A real audit finding from the demo
The Wonka audit demo grounds its factory-access finding in the engagement's approved evidence.
demo-wonka-audit/engagements/2026-factory-access/findings.md.hint ↗
# read Approved source
../../sources/approved.md
# finding factory-access-2026-1 {#auditor_demo}
Approved finding grounded in SRC-1.
# never Invented detailsWhat your assistant does next
It drafts from SRC-1, keeps the finding inside the engagement boundary, and will not infer sample results that were never observed.
What your AI assistant receives
<finding name="factory-access-2026-1">
Approved finding grounded in SRC-1.
</finding>
<never name="Invented details">…</never>Your boundaries stay visible
What it will never do
- Invent sample results, evidence or an unsupported finding.
- Assign severity outside the engagement's declared scale.
- Present out-of-scope work as an audited conclusion.
FAQ
Questions auditors ask
Does it perform the audit?
No. It carries scope, procedures and evidence context into drafting; auditors perform and evaluate the work.
Can it choose severity?
Only by applying the declared scale to documented impact and likelihood; missing inputs remain unresolved.
Will another engagement's evidence appear?
Folder scope keeps engagement-specific evidence out of unrelated working papers.
For your technical colleague
The vocabulary
A hintbook is a vocabulary for your profession—installed, not written by you.
engagement- Sets the subject, period, team and scope of the audit.
objective- States what the engagement is intended to determine.
criteria- Names the standard used to evaluate evidence.
audittest- Defines a procedure and the population or sample it covers.
evidence- Records an observed artifact and its provenance.
finding- Connects criteria, condition, cause, effect and evidence.
severity- Applies the engagement's approved rating scale.
followup- Records remediation ownership and the next review point.
Manual setup
Bootstrap is read-only: it prints instructions for the assistant. The assistant performs the installation.
npm install -g @openhint/cli
hint config
hint add @openhint/hintbook-auditor
hint apply
hint verify engagements/2026-factory-access/findings.md
hint emit engagements/2026-factory-access/findings.md