Five small tools. Evidence checks. Human decisions.
Bring a checkable trail to your AI drafts.
Security answers, test cases, interview codes, spreadsheet findings and contract changes—each with a usable local fallback and an explicit validation step.
Downloadable Python CLIs + agent Skills. Use your existing agent if you want AI interpretation; no AI subscription is included. You need to be comfortable running Python.
Select approved answers. Cite the source. Leave unsupported claims blank.
$19 USD, once
A citation is not proof that a policy is true or relevant. A security owner must review every answer. No compliance certification or native SIG/CAIQ Excel importer.
Draft requirement-linked cases. Gate missing Must coverage.
$12 USD, once
100% ID coverage is not semantic coverage or a passing application test. Offline mode does not discover edge cases. Assumptions and negative/boundary expectations need QA review.
Code interviews with exact quotes you can trace back to the source.
$12 USD, once
Exact quotes do not prove a sound interpretation, independent sampling or population prevalence. Only codebook mode was live-tested. No audio transcription; use consented/deidentified text.
A local static first pass for formula risks—not a correctness certificate.
$15 USD, once
Static audit only: no formula calculation, automatic fixes, financial assurance or complete correctness check. Named/dynamic/structured/whole-column references are flagged, not resolved; ranges cap at 10,000 cells. Heuristics may misfire.
Connect a text diff to your playbook. Keep legal judgment with a reviewer.
$15 USD, once
Not legal advice. A qualified reviewer must inspect the originals. Keyword risk is not legal analysis. DOCX headers/footnotes/comments/tracked revisions are not fully supported. PDF requires optional pdftotext, may lose layout and has no OCR.