Customer Security Reviews for Software Vendors: Secure Development and Change Control Evidence for SIG Lite, CAIQ, and Enterprise Vendor Risk Questionnaires
The deal is approved. Then the security questionnaire arrives.
The SDLC section is where it stalls.
Every enterprise sale ends in a vendor security review: a SIG Lite, a CAIQ, or a customer’s own 200-question spreadsheet, followed by requests to back up the answers. The infrastructure questions have a SOC 2 report behind them. The secure development questions usually have a paragraph and hope. SDLC Playbook produces the engineering evidence that turns the paragraph into a package, continuously, from GitHub, Azure DevOps, Jira, Slack, and Teams.
“Describe your secure SDLC” is easy. “Provide evidence for the last quarter” is not.
Enterprise vendor-risk teams have standardized on a handful of instruments: Shared Assessments’ SIG Lite and SIG Core, the Cloud Security Alliance’s CAIQ, and in-house questionnaires that borrow from both. All of them carry a section on application security and secure development: do you require code review, do you separate duties on release, do you test before deploying, do you track changes to tickets, how do you handle emergency changes, how is AI-generated code governed. The answers are yes. The follow-up, increasingly, is show us.
That is where deals slip a quarter. The SOC 2 report covers the controls the auditor sampled, for the period it covers, and says little about the development process in the detail a reviewer now wants. So a senior engineer stops shipping and assembles screenshots of pull requests, branch-protection settings, and pipeline runs, a different set for each customer, none of it verifiable, all of it stale by the next renewal.
The questionnaire is not the blocker. The evidence behind the SDLC answers is.
The questionnaire sections that reach into the development process.
A starting-point map, not a compliance or legal opinion. Your assessor’s, examiner’s, or customer’s interpretation governs, and scope varies by organization.
What a vendor-risk reviewer asks for behind the SDLC answers.
Use this as the list to test your own answers against before a customer does. For each production change during the period the reviewer names, can you produce the following without reconstructing it?
- A linked work item that authorized the change, with acceptance criteria written before the work started.
- A code review by someone other than the author, with the reviewer identified and the approval timestamped.
- Test results for the change, including coverage on the changed code and any required test plan.
- Security and dependency scan results from your CI pipeline, with high-severity findings resolved or formally accepted.
- A release approval by an authorized approver who is not the author or the deployer.
- A rollback plan attached before deployment, not written after.
- The deployment record: who deployed, when, from which build.
- Emergency change handling for any skipped gate: justification, approver, follow-up task, and its resolution.
- AI involvement: which changes were AI-assisted, and who reviewed them.
- Integrity and period coverage: a way to show the evidence has not been edited since capture, and the same evidence for every change in the window, not only the ones the reviewer picks.
Every item on that list, captured when the control runs.
Gates that produce evidence
Six PR gates run by Code Sentinel (ticket linkage, non-author review, coverage, and more) and six release gates run by Release Gatekeeper (approval, rollback plan, test signoff). Each gate result is stored as evidence mapped to the control it satisfies.
Separation of duties, enforced
A Release Approver role with deny-by-default deploy gating. Nobody approves their own release. Approvals given in Slack or Microsoft Teams are captured as control-mapped evidence at the moment they happen.
Overrides with a paper trail
When a gate has to be bypassed, the override records justification, approver, follow-up task, and audit tag. The exception becomes evidence instead of a gap.
Tamper-evident vault and export
Every evidence item carries a canonical-JSON SHA-256 and a verify endpoint confirms it has not changed. One click produces a PDF audit package and an evidence ZIP with a chain-of-custody manifest, scoped to the period you were asked about.
Drift detection across the period
Governance Drift Detection reconciles the process you documented against what actually ran. Each control reports CONFIRMED, OBSERVED_ONLY, or NOT_OBSERVED_IN_WINDOW, so you find the gap before a reviewer samples it.
AI-assisted code, accounted for
Commits and pull requests carrying AI co-author or bot signals are recorded, so when a reviewer asks which changes were AI-assisted and who reviewed them, the answer comes from evidence rather than recollection. Where no signal exists the product reports unknown, never human-only.
Integrations shipped today: GitHub, Azure DevOps, Jira, Slack, and Microsoft Teams (commercial cloud). Evidence is mapped to SOC 2, NIST SSDF (SP 800-218), and NIST 800-171 today; one finding lights up every framework the control belongs to. Direct mapping of evidence to SIG or CAIQ question identifiers, and pre-filling questionnaire answers from evidence, are on the roadmap; today you attach the exported package to the answer.
What SDLC Playbook does not do.
It does not fill out your questionnaire, and it does not produce a SOC 2 report. It produces and maintains the engineering evidence that backs the secure-development answers, collected continuously and stored tamper-evident, exported for the period a reviewer names.
It does not replace a GRC or trust-page platform. Vanta, Drata, SafeBase, and their peers manage the questionnaire workflow, policies, and the infrastructure-side evidence. SDLC Playbook covers the development-process slice they cannot see; many teams will use both. See the Vanta comparison for where the line sits.
It does not yet map evidence to SIG or CAIQ question identifiers or pre-fill answers; both are on the roadmap. SDLC Playbook itself has not completed a SOC 2 examination; our security page says exactly what we hold and where we stand.
What software vendors facing enterprise security reviews ask, answered.
What do SIG Lite and CAIQ ask about secure software development?
Both ask whether you have secure coding standards, review code before release, test and approve changes, separate development from production, control production changes, and handle emergency changes with documented exceptions. Enterprise buyers increasingly add a question on how AI-generated code is reviewed and tracked.
Our SOC 2 report covers change management. Why do reviewers ask for more?
A SOC 2 report describes controls and the auditor’s sample for a past period. It gets a vendor onto the shortlist. What keeps the vendor on it is whether, when the customer’s security team asks who approved one specific change and what testing ran before it shipped, the answer comes from a system of record or from chat threads and memory. Reviewers now want current, verifiable evidence that the development process operates, often for a period the report does not cover, and in more detail than the report provides on code review, testing, and exceptions.
Does SDLC Playbook fill in the questionnaire for us?
Not today. It produces the evidence package behind the secure-development answers, scoped to the period the reviewer asks about. Mapping evidence to SIG and CAIQ question identifiers and pre-filling answers from evidence are on the roadmap.
How does this shorten the review?
Reviewers stall on unverifiable screenshots and on gaps they find in the sample. Evidence captured at the moment each gate ran, hashed, and exported with a chain-of-custody manifest answers the follow-up in one attachment, and drift detection shows you the gaps before the reviewer does.
Can it show which of our changes were AI-assisted?
It records commits and pull requests that carry AI co-author or bot signals and reports them per repository with the reviewer on each. Where no signal exists it reports unknown rather than human-only.
Enterprise deal waiting on a security review?
Send the evidence package instead of the screenshots.
The product is built and in private design-partner stage. Design partners get 90 days free and early-partner pricing in exchange for reference rights and honest feedback. If the SDLC section of every questionnaire costs you an engineer-week, let’s talk.