Runnable proof that the field kit can become a validated handoff.
This is a cleanroom reference implementation: structured JSON goes through required-field validation and renders into a markdown deployment receipt, product-feedback packet, and post-workshop handoff with friction deltas, safety gates, model behavior, tool gaps, adoption notes, and ready-state lineage across a second service surface. The latest path aggregates those fixtures into a simulated public-safe product feedback export with routing, severity, release gates, and promotion rules, then scores the field-trial lane against role needs and no-outbound gates.
python3 tools/deployment_receipt.py validate --input examples/deployment_receipt_input.json
python3 tools/deployment_receipt.py validate --input examples/deployment_receipt_repair_input.json
python3 tools/deployment_receipt.py validate --input examples/deployment_receipt_reject_input.json
python3 tools/deployment_receipt.py render --input examples/deployment_receipt_input.json
python3 tools/deployment_receipt.py feedback --input examples/deployment_receipt_input.json
python3 tools/deployment_receipt.py validate-handoff --input examples/post_workshop_handoff_input.json
python3 tools/deployment_receipt.py handoff --input examples/post_workshop_handoff_input.json
python3 tools/deployment_receipt.py validate-handoff --input examples/post_workshop_handoff_ready_input.json
python3 tools/deployment_receipt.py handoff --input examples/post_workshop_handoff_ready_input.json
python3 tools/deployment_receipt.py validate-feedback-export --input examples/product_feedback_export_input.json
python3 tools/deployment_receipt.py feedback-export --input examples/product_feedback_export_input.json
python3 tools/deployment_receipt.py validate-scorecard --input examples/field_trial_acceptance_scorecard.json
python3 tools/deployment_receipt.py scorecard --input examples/field_trial_acceptance_scorecard.json
python3 -m unittest tests/test_deployment_receipt.py tests/test_proof_ledger.pyWhat exists
What the tool rejects
- Scope gaps Missing owner, task, verifier, data boundary, rollback path, human review rule, or decision owner.
- Receipt gaps Missing context loaded, files changed, command result, security boundary, decision, or known limits.
- Decision drift Any decision outside ready, repair, or reject. All three states are represented as fixture files.
- Feedback gaps Missing ready patterns, trust failures, product implications, or decision-specific follow-up.
- Handoff gaps Missing SDLC friction deltas, safety gates, model/tool feedback buckets, next owner, next check, or handoff assets.
- Export gaps Missing simulated/public-safe source boundary, routed signal buckets, severity, release gates, promotion rules, or no-customer-claim language.
- Scorecard gaps Missing role-need coverage, low average score, blocked ready-state gates, private evidence paths, non-official role URLs, or loss of the 2026-05-26 no-outbound gate.
Why this matters
The role asks for demos, reference implementations, workflow automations, technical content, product feedback, safety, and operational readiness. This reference turns the receipt pattern into runnable source, turns the same validated run into a product-feedback packet, and adds a second handoff path for the end of a workshop. The handoff path now shows a repair-state first pass and a ready-state repeat on another public-safe service surface, then routes the fixture lineage into a simulated product-feedback export without pretending it is real customer rollout evidence. The scorecard decides whether that fixture-backed field lane is ready to show.
This is not real customer rollout evidence. It is a code-backed operating proof: a demo should produce structured evidence and a workshop should leave behind a friction delta a second engineering leader can validate without the presenter in the room.
Inspect next
Cleanroom line
This page is independent from OpenAI. It uses OpenAI and Codex only to name public role context. It does not use OpenAI logos, product UI, private systems, or affiliation language.
Sources checked: AI Deployment Engineer - Codex and OpenAI interview guide.