Chara cycles and D60: implementation validation receipt

This records the earlier sign-only checkpoint. The subsequent full-profile extension admits a separately named modern profile; classical continuous-degree attribution is still unestablished. Counts and decisions below remain this checkpoint's record.

Date: 7 October 2026. Decision: approved plan slices A/B/C complete locally; slice D completes the sign-only source admission branch. Source-owned continuous D60 placement remains explicitly open. VED-001 is locally complete; VED-003 is a bounded admission with that remaining subitem.

The approved plan, Chara standard and D60 source standard govern scope, source locators, compatibility and remaining evidence. The user authorized implementation, not Git publication, release or deployment.

Runtime and scope

• Checkout C:\dev\moira, branch main, starting commit 67b66de14cde0447730b0c9b57e5e14c1edd9935; Moira 6.9.9.
• Project .venv\Scripts\python.exe, Python 3.14.3; required native extension moira/_moira_native.cp314-win_amd64.pyd loaded. These doctrine/transport changes have no admitted native counterpart and require no speculative port or native rebuild.
• MOIRA_TEST_MODE=1, MOIRA_STRICT_KNOWN_ISSUES=1, MOIRA_NO_DOWNLOAD=1; empty known-issues register, network deny by default, marked loopback only, external-network tests excluded.
• Resolver selected installed C:\Users\nilad\.moira\kernels\de441.bsp. Harness content probes confirmed DE441. Three new chart tests use the shared moira_engine serving reader and actual sidereal chart context; separate stubbed-context tests isolate transport/preflight. No kernel downloads.
• Initial modified work register and untracked research/plan documents were preserved. No unrelated engine family, website/Urania, kernel, harness, known-issue exemption, golden/snapshot artifact or tolerance was changed.

File group	Result
jaimini_extended, Chara server models/route	Strict cycle/body/node/finite preflight, representable ordered intervals, frozen actual computation receipt; existing arithmetic retained
varga, Vedic facade	Four source-name corrections, explicit D60 enum/sign result/helper, canonical method receipt, sign-only strength selection and full-point rejection
Varga models/route/service/serializer	All eight placement shapes preserve nullable fields; new typed sign route; Vimshopaka receipt; truthful method applicability in OpenAPI
Root/facade/Vedic curation	Shared identities for CharaDashaComputation, D60Method, D60SignResult, d60_sign; new Moira.d60_sign admitted in method list and machine contract
Tests	New source/boundary/preflight/transport/strength/real-chart cases and explicit public API admission declarations
Standards/API/REST/Home/changelog/register	Current executable contract, migration values, source limits, generated REST inventory and dated completion links

Evidence and compatibility

Rao's published second-cycle Aries/Taurus durations are partial support for Moira repetition. Synthetic fixtures isolate those durations and do not claim to reconstruct the published chart/dates. Every retained Scorpio/Aquarius co-lord branch is tested as implementation truth; complete source-chain collation remains open. Fixed-year, selected lord mode, actual count, direction and continuous ordered mahadasha/antardasha boundaries are checked. Legacy five-field result construction retains unknown metadata instead of inventing a historical convention. Invalid engine and HTTP inputs fail explicitly before doctrine dispatch.

Santhanam BPHS I printed p.83/PDF p.82 was visually checked against its fingerprinted local file. An additional primary-text transcription was collated; its unknown printed base edition and Dandabhrt/Dandayudha variant are recorded in the D60 standard. The commentary supports a sign law and the Capricorn-to-Pisces example. Tests derive other signs from that law across all twelve natal signs, both parities, exact half-degree boundaries and circular limits. They supply no missing continuous-degree authority.

Literal odd/even ordinal fixtures cover the four corrected names: 6 Kinnara, 37 Sudha, 52 Dandayudha, 59 Bhramana. Repeated names remain legal. Shared serializer tests cover all eight placement routes, including generic D60 deity=null; three actual DE441 chart requests confirm placement transport and provenance. These are composition/transport checks, not an external astronomical-accuracy or predictive-validity oracle.

An executable comparison against the trusted starting-commit Python modules verified 720 D60 midpoints, 48 Chara cases (all Lagna signs, both cycle counts, with/without nodes) and 28 default Vimshopaka cases (seven planets, four groups). Default numerical fields/periods/strength entries and totals agree exactly. Exactly 48 deity occurrences change: four corrected ordinals across twelve signs. Textual source claims and additive metadata/JSON fields change intentionally. Legacy/manual full points and strength results may keep unknown d60_method=None.

Independently calculated strength consequence

Synthetic input: Sun 130, Moon 200, Mars 125, Mercury 315, Jupiter 105, Venus Capricorn 13 degrees 25 minutes, Saturn 190 (sidereal degrees). The existing D1 friendship convention makes Mercury Venus's great friend and Jupiter its enemy: vargavishwa 18 versus 7. D60 sign/lord changes from harmonic Gemini/Mercury to source Pisces/Jupiter. This is a convention-owned arithmetic fixture, not a new independent validation of the friendship table.

Group	D60 weight	Harmonic D60 points	Source D60 points	Harmonic total	Source total	Delta
Dashavarga	5.0	4.5	1.75	16.575	13.825	-2.75
Shodashvarga	4.0	3.6	1.4	16.025	13.825	-2.2

Every other division entry is unchanged; delta equals (7-18)*weight/20. HTTP carries the same selected/applied method and contribution. Groups without D60 return null applied metadata and reject a nondefault selection.

Commands and results

All commands used the project runtime and the environment above. Private machine-readable output resides in C:\dev\outputs\vedic-001-003-research-2026-10-07; it is not a publication input.

Selection	Result / receipt
Chara, deity, Varga, Shodashvarga, new admission, Vedic exports/facade, public audit, existing/new Varga HTTP (ten files)	Initial broad run: 366 passed, one failed, zero skipped; implementation_core.xml. Failure was the missing explicit admission of new Moira.d60_sign in the method-name declaration
New engine/HTTP admission files plus complete public audit after correction	170 passed, zero failed/skipped; implementation_final.xml; covers every formerly failing node and final method/result/serializer code
Existing direct-input HTTP tests, -k chara	10 passed, zero failed/skipped; implementation_chara_http.xml
Existing Phase-2 HTTP tests, -k vimshopaka	3 passed, zero failed/skipped; implementation_strength_http.xml
Final Chara/D60 HTTP selection after fixed-year OpenAPI constant	54 passed, zero failed/skipped; implementation_final_http.xml
Trusted-baseline executable compatibility probe	All 796 sampled cases pass exactly; implementation_compatibility.json
Scoped Ruff comparison	Zero introduced diagnostics in fifteen changed Python files. Existing 60 E402 root-import and five F811 facade-import diagnostics match starting commit; lint_differential.json. No broad lint cleanup or suppression
Documentation, generated REST and in-memory wiki preview	Owning doc guard, REST generator/check, new local links/code fences and in-memory page rendering pass; existing changelog file-URI links are preserved, not certified; generated Git wiki is not written
Git whitespace	git diff --check passes

The focused pytest selections cover 380 distinct test nodes, all passing in their latest applicable run, zero skipped. Counts are deduplicated across reruns; no claim of a single 380-test invocation or full repository suite. The ordinary admission declarations are updated for explicitly added API; protected snapshots/goldens and test policy remain unchanged. An initial fixture-only failure (a manually allocated facade lacked its reader field) was corrected in test setup before the broad selection. One attempted command named a nonexistent standalone Vimshopaka unit file and collected no tests; the actual owning Varga unit file and HTTP selections above supply coverage.

Reproduction uses python -m pytest <listed owning files> -m 'not external_network', python scripts/check_doc_consistency.py, python scripts/sync_rest_api_reference.py --check and git diff --check, with .venv\Scripts\python.exe substituted for python. No blanket full-engine accuracy, full-suite, native-parity or all-Vedic-source claim follows.

Remaining gate and publication state

Continuous source-owned D60 placement needs an identified edition/passage defining within-sign degrees, endpoints, wrap and parity, plus source-owned numerical fixtures. The admitted sign-only route and Vimshopaka selection do not borrow harmonic degrees or close that gate. Full-cycle/co-lord Chara source collation remains a separate research frontier.

Code/tests/canonical documents are locally changed. No commit, push, version bump, release, generated-wiki publication, deployment or website/Urania adoption is included in this request. The work register preserves these separate states.