A runnable recipe, not another proof memo.

A reviewer needs to see whether the operator can turn AI coding workflow knowledge into a small, teachable, runnable example: clear problem, bounded input, steps, verifier, expected output, failure gates, and extension route.

1. problem_statementName the reader problem.The artifact teaches the proof route instead of selling it.
2. minimal_inputConstrain the input.Only public-safe fixture text, public docs, public URLs, and verifier commands are allowed.
3. recipeRender the handoff.Commands, expected markers, release gates, role signals, and reviewer notes become one repeatable packet.
4. verificationRun the public-pack verifier.The example cannot pass while missing required docs, fixtures, tests, or claim-boundary checks.
5. handoffPublish only after checks.The reviewer-facing page is connected to the graph after local verifier, privacy scan, and public URL checks pass.
6. extensionExtend only for new friction.Future examples need a new verifier, source, or customer-workflow friction point.

1. examplesConcrete example format.A runnable packet instead of only strategy copy.
2. guides_patternsGuide linkage.Connects the example to the existing deployment-pattern library.
3. best_practicesReviewable practice.Data boundary, permission, approval, rollback, and security communication are explicit.
4. reference_implementationExecutable source.The validator and renderer make the example machine-checkable.
5. customer_workflow_designWorkflow shape.Intake, workshop, receipt, review, and feedback are mapped without customer-traction claims.
6. security_operational_readinessGate discipline.The security review packet is upstream of the example.
7. product_feedbackSignal routing.Recipe observations can become product feedback without claiming real customer adoption.

• python3 tools/deployment_receipt.py validate-cookbook-example --input examples/cookbook_style_codex_example.json
• python3 tools/deployment_receipt.py cookbook-example --input examples/cookbook_style_codex_example.json
• python3 scripts/verify_public_pack.py
• python3 -m unittest tests/test_deployment_receipt.py

1. Public-safe dataRepository files, public proof URLs, and executable commands only.
2. No target-company contribution claimNot an OpenAI Cookbook submission and not affiliated with OpenAI.
3. No customer or traction claimNo customer rollout, revenue, funding, legal entity, employment, endorsement, or real-world impact claim.
4. No outbound before 2026-05-26The example is proof packaging only until the date gate, an inbound recruiter signal, or a current explicit override changes the route.