Skip to content

Build expanded English public gonol and recursive ordinal construction - #88

Open
erinepshovel-code wants to merge 25 commits into
mainfrom
feat/english-ordinal-public-gonol
Open

erinepshovel-code wants to merge 25 commits into
mainfrom
feat/english-ordinal-public-gonol

Conversation

@erinepshovel-code

@erinepshovel-code erinepshovel-code commented Oct 14, 2026 •

Copy link
Copy Markdown
Contributor

Authorization and selected construction

Operator: merge, retire the previous completed geometry, replace it with this construction. english.ordinal/1 is the supported default, without legacy fallback. Radius is R/2**(n-i), zero-based; angle comes from the declared expanded carrier. Ordered interiors, repeated occurrences, whitespace and source provenance survive. UCNS #241's exact consumed producer is a7bfa0c96daca1f7445ba57c39fa6d8fd62e51eb.

Settled answers: 1. SPACE; 2. SPACE; 3. native Mobius integration not required; 4. reuse prior definitions/functions on the new base. Question 5, vocabulary growth, remains open and does not prevent using the fixed complete corpus.

Executable definition migration

Runtime commit: 2ba7131c5931cccbd779ed08b59f8a2a356729d4.
Provenance head: 3c8e405edcb960ac80e6630609b82d62c301d885.

ordinal_definitions.py consumes the unchanged previous definition table and final-y behavior. It iterates the expanded carrier, preserves overlapping roles and number-name links, and constructs all character/definition interiors through UCNS ordinal affixiation. It never calls the historical constructor. Missing source definitions stay explicit hmmm.

The supported build requires character-definitions.json and binds its receipt in the geometry manifest. Verification independently replays its entries, source, positions and nested receipts. Missing, extra, reordered or altered components fail. Previous artifacts require explicit rebuild; historical receipts do not certify this runtime.

File plan, risks and witnesses

Paths Change and purpose Risk / required test
english_gonol/ordinal_construct.py Replace geometry and require definition component in build/verify exact-source full-corpus replay, altered/missing data refusal
english_gonol/ordinal_definitions.py (new) Preserve inherited data/behavior, reconstruct on new carriers prior-method comparison; no historical constructor; independent all-position replay
tests/test_ordinal_definitions.py (new), tests/test_ordinal_construct.py, tests/test_active_construction.py inheritance, corruption, routing and provenance regressions isolated pinned UCNS environment; fixtures are not corpus evidence
__main__.py, ACTIVE_CONSTRUCTION.json selected entrypoint, no fallback retired-selector/selection-mismatch tests; help/profile without compute
both English workflows isolated environments, definition tests, complete construction/query Unicode acquisition precedes artifact identity checks; baseline-owned sealed source closure
BASE.json, root manifests, ordinal work graph source-qualified dependencies and coherent runtime pin recomputed root digest; byte comparison to actual selected source; no canonical libs refresh
README, ordinal docs, migration witness preserve operator decisions and exact observed scope source/implementation/evidence kept distinct

All paths above are relative to research/english-gonol/ except workflows and root manifests. The source-admission helper closure is retained unchanged as support, not a competing completed geometry. Physical helper separation and downstream consumer migrations remain separate work.

Executed evidence

85 local focused checks passed: the new definition cases, current ordinal cases and exact packaged UCNS ordinal tests. The local harness poisoned the unused historical constructor import. This is scoped local evidence, not whole-repository CI.

Complete supplied-field definition replay covered 4,698 characters, 4,757 definition closures and all ten number-name interiors, using the full 103,053-word source inventory. All ten word-interior receipts matched the supplied complete geometry. The inherited source leaves 4,643 scalar definitions explicitly unresolved. See experiments/ordinal-definitions/migration-witness.json for source hashes, implementation hash, receipt and scope. This run is not a new full-corpus source-admission replay.

New root work-graph digest: a28ff770ccb5c7cdae017272dd47e3bacb0b7054b07ebe7086fe26d534aaebb6.

Current-head gates and hmmm

Current-head hosted construction/replay, applicable checks and review must be observed before a guarded merge. Earlier successful runs are historical checkpoints only. Source integration is not UCHC release, downstream deployment, semantic usefulness or measurement validation. SPACE references are settled while global transport witnesses remain engineering work; native Mobius integration is excluded. Source-uncovered definitions and vocabulary growth remain explicit, not invented.

Copy link
Copy Markdown
Contributor Author

Actual complete-corpus construction and replay

Run 38528624494 at constructor head 4bb56be202c0c559416ffcfc0be3e3c781b57423 completed successfully, including source acquisition, 40 cross-repository tests, full construction, independent verification, and source-exact query recovery.

Observed artifact results:

  • Public gonol: 4,698 unique scalars, from the exact union of 118 source-admitted scalars, 157 legacy anchors, 2,247 Unicode number scalars, 2,384 Math scalars and the 31-scalar logical supplement; these sets overlap.
  • 103,053 closed words, 420,006 frames, 4,792,299 placements.
  • All 73 pinned corpus files retained and verified; source reports 135,969 lexical entries and 107,519 synsets.
  • Query the geometry recovered exactly, including its SPACE participant and both word interiors.
  • A separate local SQL audit of the downloaded actual databases checked every placement: zero incorrect ordinal radii, ordinal sequences, carrier indexes or interior references; geometry occurrence counts exactly match the preserved source.

Actual construction artifact: 11877926403, 224,361,408 bytes; archive SHA-256 e0d6541727bc2faa9ef818e70a2a068fe35604971076b0da604e3ed0c9df38e5.
Evidence artifact: 11877517115; archive SHA-256 128f084240a3ab1e3e66d88744672dbdfa142fa6edc4ec6f5a17b4bd1e0f9ef4.
Both archives were downloaded and their hashes checked. Their manifests/public registry agree byte-for-byte.

Canonical public-registry digest: dc8740a258b95688882b70d4faaa8b96b8f903990f459f4804e779879fca043d.
Geometry receipt: 5a49c4dd9e6247ada7c5c9ff4c5f11303815fcb05b1cca03f40d7d50557c2cb8.
English implementation SHA-256: 65039f5cfa5d17dbb4a432c1cc90e0f9e55ba751dfd465d96e461a1f175796d6.

The subsequent integration head 2cf9d504680557537276d44f3ce5179ea28d2c56 changes only the existing English test workflow's dependency installation. Its constructor blob remains cdf94ef8344fca131b34912ed590f73973112785; the broader english-gonol run 38529558550 has now passed. This is not a claim that every other exact-head workflow has finished.

hmmm: Both PRs remain unmerged. Layout recipes are explicit profile choices. Native Möbius lift, global frame transport and semantic inference are not established by local construction/recovery success. Actions artifacts have 30-day retention; downloaded construction copies remain separate from that retention window.

…se unresolved laws

Operator-authorized replacement. Add a no-fallback package entrypoint and source-owned active selection, nine routing regressions, updated usage, manual full-build dispatch, and prioritized construction questions. Preserve the byte-sealed admission dependency closure and previous geometry receipts as historical support; do not label legacy outputs as new geometry. UCNS #241 merged at 6ebbf970ad13a0d2e1c13bd65c496b671e740185; consume the same exact producer bytes. Local focused regression: 49 passed. Root authority, libs pins, and UCHC release standing unchanged.
@erinepshovel-code
erinepshovel-code marked this pull request as ready for review October 14, 2026 09:29
@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Oct 14, 2026 •

Copy link
Copy Markdown

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-10-14T13:17:26.210819Z 08ccfe6 New commits
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 0e0ea9ebe6

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread research/english-gonol/ACTIVE_CONSTRUCTION.json
Comment thread research/english-gonol/english_gonol/ordinal_construct.py
Comment thread .github/workflows/english-ordinal-affixiation.yml
Comment thread research/english-gonol/english_gonol/ordinal_construct.py Outdated
…guage boundaries

Register the selected source and six exact input roles in root machine/human projections; update English BASE without rewriting extraction provenance or libs snapshots. Restore identity, recursive scale and EDCM measurement-only clauses caught by unchanged language-authority tests. Add source-registration/digest regression. Local focused rerun: 50 passed. Geometry and the thirteen-file admission closure remain unchanged.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 4c9470d90c

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread research/english-gonol/english_gonol/ordinal_construct.py Outdated
…ission CI inputs

Repair the three remaining #88 review findings. Admit only the two complete pinned Unicode 18.0.0 file hashes; synthetic parser/layout fixtures cannot be called the release. Enforce exact manifest schema and fixed uncertainty claims before replay, reject false extra status fields and malformed counts/receipts, and emit source-owned boundary text. Include the entire admission module closure in the full-corpus workflow path filter. Local focused regression: 64 passed. Complete retained UCD bytes passed the new public loader. Full hosted reconstruction is required for this changed verifier implementation; previous manifest receipts are not promoted.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 4dea37abee

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread stack-manifest.json Outdated
Comment thread .github/workflows/english-ordinal-affixiation.yml Outdated
Comment thread research/english-gonol/docs/ordinal-construction.md Outdated
Snapshot only regular nonsymlink UCD files, reject oversize inputs before open, retain at most the limit plus one byte, detect file replacement/growth, and authenticate the bounded snapshot before building or copying. Ordinal placement, public layout and SPACE-first behavior are unchanged. Runtime source checkpoint for the following atomic provenance/test/CI update; no branch ref moves at this checkpoint.
Close the four further #88 findings in one source-plus-metadata transaction. Select exact runtime checkpoint 37adccb in BASE and root projections, recompute the work graph, compare selected module bytes in CI, retain separate source-head and tested-merge evidence, and scope prior corpus claims to their original implementation. Add bounded-read/special-file/race and workflow proof regressions. Local focused suite: 73 passed. Runtime implementation SHA-256 a1933daabd2ae9323588c99f395441c7711e082ea03c97797a81051b1430aa2d. Full final-head construction and merge-tree replay required before merge; previous artifacts are not relabeled.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 174f96a724

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread .github/workflows/english-ordinal-affixiation.yml Outdated
Comment thread .github/workflows/english-ordinal-affixiation.yml
Comment thread .github/workflows/english-ordinal-affixiation.yml Outdated
…h from geometry data

Byte-bind ACTIVE_CONSTRUCTION.json and the local source work graph alongside both runtime modules to the selected source checkpoint. Trigger the full gate on BASE and root-manifest changes. Keep the work graph as explicitly separate source CI evidence, not an unverified authority record inside the construction artifact. Add three regressions and document the verification boundary. Local focused suite: 76 passed. Runtime bytes and selected source 37adccb remain unchanged; final merge-tree CI still required.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: b7f8672e15

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread .github/workflows/english-ordinal-affixiation.yml Outdated

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 69559d3c90

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread .github/workflows/english-gonol.yml Outdated
Comment thread research/english-gonol/english_gonol/ordinal_construct.py Outdated
…ents

Record operator answer 3: not required. Preserve SPACE attachment answers and outstanding witnesses; leave questions 4 and 5 open. Align README and construction guide without changing runtime, source pins, or UCNS native geometry.

Applies skill-lib gonol-build scope preservation and the-interdependency workflow.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: ee130be775

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread .github/workflows/english-ordinal-affixiation.yml
Record answer 4: reuse the previous definition method and applicable functions; reconstruct over the expanded carrier rather than redesign semantics. Bind the inspected prior definition sources, preserve overlapping roles and honest unknowns, and specify old/new preservation and no-fallback witnesses. Only vocabulary growth remains a construction question.

Documentation only: no runtime, source-pin, test, workflow or generated-artifact changes. Applies gonol-build structure preservation, the-interdependency source-first inheritance, and WDLL evidence separation.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 14341800a6

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread research/english-gonol/docs/ordinal-work-graph.json

Copy link
Copy Markdown
Contributor Author

@codex review

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 34fdda574f

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread .github/workflows/english-ordinal-affixiation.yml
Comment thread research/english-gonol/BASE.json Outdated
Comment thread .github/workflows/english-ordinal-affixiation.yml
Apply operator answer 4 in executable build and independent verification: preserve the unchanged definition table and final-y behavior, expand carrier coverage, and require source-bound character-definitions.json with no historical geometry fallback.

Add 19 definition/integration cases; update manifest fixture and isolated CI coverage. Preserve concurrent baseline-owned admission and package-isolation repairs. Acquire Unicode before verifying artifact identities.

Local focused suite: 85 passed with the unused historical constructor import poisoned. Complete supplied-field definition replay: 4698 characters, 4757 definition closures, ten number-name interiors matching corpus word receipts, 4643 explicit unknowns. This is not current-head hosted/full-admission evidence.

Applies wdll, gonol-build, msdmd, meta-module-build and stack-update. Source repinning follows in the same guarded branch transaction.
Pin active runtime to 2ba7131 in BASE and both root projections. Recompute the work-graph digest and correct stale native Mobius/operator boundaries. Preserve canonical libs pins and all unrelated participants.

The parent contains executable definition migration and local witnesses; this provenance-only commit adds no hosted CI or full-corpus certification.
Address PR #88 review 4270796319: every declared participant must map exactly once to an acquired input or the baseline whose sealed closure is checked. Reject unknown, repinned, duplicate and omitted identities; explicitly verify the fetched Stack baseline commit.

Five local positive/adversarial cases passed against the exact participant-loop excerpt; committed tests extract the real workflow loop in repository CI. No runtime, selection, work-graph or canonical source-pin bytes change. Complete current-head hosted evidence remains required.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 08ccfe65a3

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment on lines +109 to +110
assert keyed['ordinal-unicode-data']['artifact_sha256'] == implementation.UCD_RELEASE_SHA256['UnicodeData.txt']
assert keyed['ordinal-unicode-math']['artifact_sha256'] == implementation.UCD_RELEASE_SHA256['DerivedCoreProperties.txt']

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P1 Badge Reference the imported ordinal module in the provenance test

Whenever either English workflow runs test_active_inputs_are_registered_in_root_provenance, this assertion raises NameError because the file imports the module as ordinal_construct, but implementation is never defined. The provenance test therefore always fails instead of validating the Unicode identities; use the imported module name or introduce the intended alias.

Useful? React with 👍 / 👎.

Comment on lines +175 to +176
assert set(inputs) == {'english_gonol/ordinal_construct.py', 'english_gonol/__main__.py',
'ACTIVE_CONSTRUCTION.json', 'docs/ordinal-work-graph.json'}

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P1 Badge Update the expected selected-source inventory

The workflow's relative tuple now contains seven entries, including ordinal_definitions.py and both inherited-definition inputs, while this assertion still expects the previous four-entry set. Consequently, test_selection_and_source_graph_are_exact_pinned_inputs fails on every run of both English workflows even after the preceding NameError is fixed; update this expected set to match the newly sealed inventory.

Useful? React with 👍 / 👎.

manifest = _json_read(output / 'manifest.json')
_validate_manifest(manifest)
from english_gonol.structural_construct import _verified_connection, open_readonly, _schema
with _verified_connection(output / 'admission') as verified:

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Bound the nested admission manifest read

Fresh evidence beyond the earlier bounded-JSON fix is that verifying an unpacked artifact enters _verified_connection here, and that helper still reads output/admission/manifest.json with an unrestricted read_bytes() in structural_construct.py. If this nested manifest is a large regular file, FIFO, device, or symlink, verify() can still exhaust memory or block before validation; route this nested manifest through the same bounded, nonsymlink reader used for the top-level JSON artifacts.

Useful? React with 👍 / 👎.

Comment on lines +135 to +139
for relative in ('english_gonol/ordinal_construct.py','english_gonol/__main__.py',
'english_gonol/ordinal_definitions.py',
'english_gonol/language/character_definitions.py',
'english_gonol/language/data/character_definitions_v1.json',
'ACTIVE_CONSTRUCTION.json','docs/ordinal-work-graph.json'):

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Pin the definition loader's complete import closure

Fresh evidence after the earlier admission-closure fix is that the new ordinal_definitions.py imports language.character_definitions, which imports english_gonol/gonol.py, but this selected-source list omits gonol.py and the baseline IMPLEMENTATION_FILES inventory omits it as well. A PR can therefore change that import-time dependency, run altered bytes in the full build, and still pass the exact-source comparison while the graph claims the pinned ordinal-source identity; include the complete new import closure and repin the projections.

AGENTS.md reference: AGENTS.md:L72-L75

Useful? React with 👍 / 👎.

Comment on lines +15 to +16
- 'research/english-gonol/docs/ordinal-*'
- '.github/workflows/english-ordinal-affixiation.yml'

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Gate edits to the committed migration witness

When a PR changes only experiments/ordinal-definitions/migration-witness.json, this full workflow is not triggered because the path filters cover ordinal docs, tests, and implementation files but not the committed witness. The general English workflow does run, but none of its fixture-based ordinal tests recomputes the witness's full-field counts, receipts, source hashes, or SURVIVED status, so those evidence claims can drift without either the selected-source check or a complete definition replay; watch this file and compare it against the reconstruction produced from the pinned inputs.

Useful? React with 👍 / 👎.

Comment on lines +100 to +102
graph=json.loads(pathlib.Path('research/english-gonol/docs/ordinal-work-graph.json').read_text())
body={k:graph[k] for k in ('repositories','boundaries')}
assert hashlib.sha256(json.dumps(body,sort_keys=True,separators=(',',':')).encode()).hexdigest()==graph['work_graph_sha256']

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Validate the source graph's fixed boundary claims

When a provenance repin changes ordinal-work-graph.json, this check proves only that its repositories and boundaries reproduce its own digest; it never validates the graph schema/version or requires the non-transfer fields to remain false. A repin can therefore set authority_transfer or proof_status_transfer to true, recompute the digest, and pass this job while the root projections still declare non-transfer boundaries; validate the exact graph shape and reconcile its fixed boundaries with the root manifest before retaining it as evidence.

AGENTS.md reference: AGENTS.md:L72-L75

Useful? React with 👍 / 👎.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant