This repository was archived by the owner on May 13, 2026. It is now read-only.
fix(org): dual-read fallback for pre-tag molecules (alpha BLOCKER 3e063) - #564
Merged
Conversation
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
Pre-existing molecules became unqueryable after `POST /api/schema/{name}/
set-org-hash`. Writes new and reads new both went through the org-prefixed
keyspace, but atoms and molecule refs stored before the tag still lived at
the unprefixed (personal) keys, so `refresh_from_db` and
`fetch_atoms_with_key_metadata_async_with_org` came back with
`InvalidField("Atom … not found for key …")` on the same data the user
had just ingested.
Design (`docs/designs/org_shared_sync.md`) is explicit that set-org-hash
does not rewrite pre-existing keys, but the product was silent and
returned a cryptic error rather than continuing to serve the data.
Fix: dual-read at each read site — try the org-prefixed key first, and on
miss fall back to the unprefixed base key. Applied at:
- `AtomStore::get_atom_by_uuid` — the direct atom lookup used by
mutation-event replay and other singleton reads.
- `AtomStore::get_mutation_events` — scans history prefix; if the org-
prefixed scan is empty, retries the unprefixed scan.
- `FieldBase::refresh_from_db` — loads the molecule ref during query;
now falls back to the unprefixed ref key when the field has an
`org_hash` but the prefixed key is absent.
- `fetch_atoms_with_key_metadata_async_with_org` — batch atom fetch
used by SingleField/RangeField/HashRangeField `resolve_value`.
Writes are unchanged — new writes continue to land in the org prefix.
Content-addressed atoms with identical content will coexist under both
keys after mix-use (harmless; the read path prefers the org-prefixed
version).
Regression test: `test_pre_tag_molecules_remain_queryable_after_set_
org_hash` in `tests/org_key_prefixing_test.rs` — register personal
schema, ingest two molecules, promote to org schema, assert both
pre-tag molecules resolve AND a post-tag molecule resolves in the same
query, and assert the Sled tree holds both prefixed and unprefixed atom
keys (confirming the fix is read-only, no migration).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
CI Rust Format step flagged the short match arm as wrappable. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
shiba4life
force-pushed
the
fix/org-tag-dual-read
branch
from
April 20, 2026 02:45
60d1167 to
4ac5fb4
Compare
… cleartext-logging)
shiba4life
disabled auto-merge
April 20, 2026 03:07
shiba4life
added this pull request to the merge queue
Apr 20, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Alpha BLOCKER 3e063 (dogfood run 4): tagging a schema with
set-org-hashorphaned every molecule that already existed on the schema. Writes went to the new org prefix, reads went to the new org prefix, but the pre-tag atoms + molecule refs lived at the unprefixed (personal) keys. Queries returnedInvalidField("Atom … not found for key …")on data the user had just ingested.Design (
docs/designs/org_shared_sync.md) is explicit that tagging does not re-key pre-existing molecules — but the product was silent and returned a cryptic error rather than continuing to serve the data. This PR lands the dual-read path recommended in the bug report: on miss at the org-prefixed key, fall back to the unprefixed base key. Writes are unchanged; a future change can re-key old data and retire the fallback.AtomStore::get_atom_by_uuid— direct atom lookupsAtomStore::get_mutation_events— history-prefix scansFieldBase::refresh_from_db— molecule ref loads during queryfetch_atoms_with_key_metadata_async_with_org— batch atom fetch used by all field variants'resolve_valueRegression test
test_pre_tag_molecules_remain_queryable_after_set_org_hashintests/org_key_prefixing_test.rs:alpha body,beta body).org_hash+trust_domain+ propagate to every runtime field (mirroringfold_db_node::OperationProcessor::set_schema_org_hash).main).{org_hash}:atom:…keys (post-tag), confirming the fix is read-only and no migration happened.Test plan
cargo clippy --workspace --all-targets -- -D warningscargo test --workspace --all-targets(full fold_db suite green, incl. all 10 tests inorg_key_prefixing_test.rs)/qa-folddb --dogfood(run 5)Follow-ups (out of scope)
249d8— extend/qa-folddb --dogfoodharness to exercise theset-org-hashpath so this class of regression is caught automatically.Refs: gbrain
projects/alpha-e2e-dogfood-run-4,design/org_shared_sync,projects/alpha-org-schema-tagging-gap.🤖 Generated with Claude Code