8.6 KiB
Detailed full-plan supplements — current context and precedence
These supplements preserve the full text structure of 15 canonical planning/testing documents, not just their requirement IDs or summaries. Every original line belongs to a mapped section. Changes are limited to privacy generalization, included-document references and a quoted shell placeholder; no original task step, dependency, negative test or acceptance paragraph was dropped. Coverage reports every section and source line range. Private source copies and value-free transformation receipts remain with the collector, not in this proposal.
What this delivery does and does not prove
The eight earlier documents remain navigation/current-process summaries. These supplements provide detailed completion phases, all T00–T29 task instructions, test families, guarded installed/operator procedures, shared contracts, UX/guest-security/Mesh requirements, and the non-IndeeHub original plan, reconciled 35-ID/eight-app scope and detailed UAT history. The 48 requirement IDs and 30 task definitions overlap; they are not 78 independent acceptance units.
This is plan-content coverage, not proof that all scenarios ran, source is safe to publish, or release is ready. Historic evidence stays attached to its original source/checkpoint. Owner factual review of the newer summaries was bounded; it is not approval of every historical statement here. Source-owner/testing review of this full transcription and section map remains pending. Public implementation visibility and full outgoing-history security review remain separate from docs publication.
Current precedence and corrections (2026-10-24 08:34 UTC snapshot)
- Development continues on dependency-ready work. Historical blanket audit/UX holds, “Phase 0 only,” and waiting-for-routine-continue language below do not reimpose a global hold. Current authority and reviewed candidate-specific safety gates prevail; source work does not wait on an unrelated heavy slot.
- Scoped docs/WIP publication authority is separate from main merge, signing, release, deployment, financial operations, public media/events, credentials, hardware/RF and destructive changes. No embedded historical grant authorizes a new action. Read actual current repository instructions and approved source-bound controls before execution.
- One reviewed bounded sequence can encompass compile/artifact/primary/collateral/cleanup; per-phase manager chat is not an operator requirement. Exact phase caps, provenance, fresh capacity/exclusion, failure stops and two-failed-cycle escalation remain mandatory.
- B10 notification artwork is last in feature work after other lanes. Actual-device acceptance is required for full closure, not a new blanket prerequisite that stops unrelated source implementation.
- Pagination scenario means 251 items across a 250-item page boundary, not 251 pages. Latest r3 emitted 11 client records: three PUT pre-header timeouts and all five later exact-byte/ETag GETs; correlation remained incomplete. Original operation 1/outer 90 stayed failed. Subsequent exact-owned removal succeeded and owned absence was proven, but supplemental outer 1 remained failed due pre-existing unknown foreign inventory drift. No latency-cause or pagination qualification credit.
- Native frozen source
20cbb722755581693546a379b2692450a04a8e67: 28 new tests UNRUN. App signed-admission source04f9daadd3ea582f7351c9dfd413eeb593d43087: isolated 127 tests/seven suites, zero skips, raw/outer 0. Native first sequence independently NOT READY R1–R5; successor submitted for independent static review with 25 bindings and 58 fake model cases. Review found a launch budget mismatch (supervisor requests 3300 seconds while its relay rejects more than 1500), so it is not executable as submitted; review continues and real lifecycle qualification remains absent. These current facts do not upgrade older fixtures or source histories. - App source
9194e2d4a26ae7879188547cc768399c89f28cfdis owner-reported unpublished by its lane at the recorded check. Clean detached/native refs are not publication/security clearance. Sensitive-history findings block raw implementation ancestry publication; no confidential values are reproduced here. - Older timing/model allowances and dates are not current forecasts. The mandatory timing contract (S15) governs: comparable full-agent execution, failures and unknown waits; no test-count credit or invented countdown. Historical performance targets remain proposals until measured/agreed.
- Historical lower-level installed passes and withdrawn operator-visible results remain historical. Whole-product/UAT acceptance is open. Never use an old row marked PASS to certify a changed artifact.
Included originals and public destinations
| Source label | Original document basename | Detailed public document |
|---|---|---|
| S01 | Archipelago-1.9.2-completion-execution-plan.md |
01-Completion-Execution-Plan.md |
| S02 | START-HERE.md |
02-Execution-Entry.md |
| S03 | TASKS.md |
03-Tasks-T00-T29.md |
| S04 | TESTS.md |
04-Qualification-Test-Contract.md |
| S05 | YAYA-UAT.md |
05-Guarded-Installed-UAT.md |
| S06 | APPROVED-UX-AND-ADDITIONS.md |
06-Approved-UX-and-Additions.md |
| S07 | CONTRACTS.md |
07-Shared-Contracts.md |
| S08 | UAT-CLOSURE-CHECKLIST.md |
08-UAT-Closure-Checklist.md |
| S09 | PUBLIC-GUEST-INDEEHUB-SECURITY.md |
09-Public-Guest-Security.md |
| S10 | ZAPS-AND-FILM-DETAILS-UX.md |
10-Film-Details-and-Zaps.md |
| S11 | MESH-FOLLOWUP.md |
11-Mesh-Investigation-and-Acceptance.md |
| S12 | TAKEOVER.md |
12-Non-IndeeHub-Original-Plan.md |
| S13 | PLAN-SCOPE-RECONCILIATION.md |
13-Non-IndeeHub-Scope-Reconciliation.md |
| S14 | UAT-TICKLIST-20261022.md |
14-Non-IndeeHub-Detailed-UAT.md |
| S15 | TIMING-RULE.md |
15-Timing-and-Estimate-Contract.md |
Referenced materials and scope boundaries
Repository source-code paths and named repository contracts are retained as references, not silently bundled or asserted reviewed. The master retained-regression checklist and other repository implementation contracts must be read at the exact selected source revision before execution; these 15 documents explicitly require that reconciliation. Private evidence labels identify retained proof categories, not public download links. Do not fabricate a public URL or present their absence as passing evidence.
The machine-readable private ledger is not copied: its fixed requirements/task map is in the earlier Test-Matrix and the full task document here. Raw status flags, wallet/session data and operational records are not publication artifacts. The private budget/usage audit is excluded because it contains private accounting and superseded blanket controls; its enduring two-cycle/provenance/resource/reuse requirements are fully retained in the included plan, test contract, timing rule and current overlay. No accounting values are needed to reproduce an acceptance criterion.
This coverage claim is limited to the enumerated canonical plan/test set. It does not assert that every referenced repository specification, every raw receipt or all historical research was transcribed. Any additional canonical requirement-bearing document identified by the owners must be added and mapped rather than silently treated covered.
Privacy and runnable-command boundary
Private home/checkout/evidence locations become generic role or evidence labels; runtime identities are removed; machine/model aliases are generalized. Repository code paths, public protocol concepts, required test assertions and source/artifact bindings remain. Examples containing placeholders are not executable instructions until specialized to current reviewed tooling. Rejected/unreviewed control scripts are not published as runnable recipes. Secret-free isolation, no sensitive mounts/live wallets, explicit lifecycle ownership and fresh execution authority remain prerequisites.
No source content was omitted for convenience. Redacted information is access/identity/location metadata, not task semantics. Reviewers must inspect the full text and transformation categories before publication; a narrow scanner is not a security guarantee.