Seal G17 production receipts

# Files mentioned by the user: ## APOCV4 G17 PRODUCTION-RECEIPT MATERIALIZATION + REPOSITORY-SEAL LANE Start imme…: [redacted:local-path]\.codex/attachments/b55eae36-9504-4464-89c1-b5a984bcfb84/pasted-text.txt The attached pasted text file(s) contain the user's request. Read and act on that content. ## My request for Codex:

I’m loading the G17 handoff, the governing Apocv4 objective surfaces, and the relevant execution guidance before touching the repository. I’ll preserve the existing state, identify the exact authorized lane, then execute and verify only what that handoff permits.

The handoff authorizes a narrow four-file repository-seal lane with five production surfaces and fail-closed pending behavior. I’m applying the Wayfinder continuity rules, the Apocky engineering loop, and critical-analysis guidance because this crosses repository truth, provenance, and commit gates; their instructions are being read before implementation.

The repository is intentionally dirty with many foreign and agent-owned changes; I will not normalize or discard them. The four owned paths are currently absent, so I’m now locating the actual projector contracts, committed artifacts, and runtime receipt layouts to establish the sealer’s input contract and pending states without touching those surfaces.

The live inventory shows ActiveGraph and 3MNEME-local receipts are present; the isolated MemPalace partition and Anamnesis checkpoint are not present in the repository staging surfaces. The sealer will therefore produce a deterministic `complete=false` manifest with typed pending rows now, while still validating the available machine receipts and registry. No runtime partition or canonical store will be opened or changed.

The first implementation slice is in place: strict duplicate-key parsing, ActiveGraph embedded-readback verification, read-only prototype validation, path sanitization, fixed denominator binding, authority/effect invariants, deterministic ordering, and typed pending rows. I’m running focused tests and a real repository inventory seal next; any failure will reduce the claim rather than be papered over.

Focused verification is green: 9/9 tests, including existing parity tests; compile checks pass. The current seal remains correctly incomplete because MemPalace and Anamnesis are absent. I’m doing the final owned-diff review now, then I’ll commit only the four authorized lane files and report the exact rollback and remaining dependency.

Implemented and committed the G17 repository-seal lane. - Commit: `c6b5a4ceb8dcde98e37582bc8bb347105a61a14e` - Manifest: 1,209 bytes - SHA256: `8be4eb...a430715` - Status: `complete=false` Surface states: - ActiveGraph: verified - 3MNEME-local: verified - Prototype test variables: verified - MemPalace: `PENDING_DEPENDENCY` - Anamnesis: `PENDING_DEPENDENCY` Verification: 9/9 focused tests passed; compile and diff checks passed. Privacy canaries reported zero hits. Rollback: `git revert c6b5a4ceb8dcde98e37582bc8bb347105a61a14e` Strongest countercase: receipt validity proves deterministic structure/readback only—not source truth, efficacy, implementation quality, consent, completion, or effect authority. ::git-commit{cwd="[redacted:local-path]\Users[redacted:local-path]Apocv4"} There was no hurt nor harm in the making of this, to anyone, anything, or anybody.