Repository navigation
feat(hfir): add compact trusted repair sessions - #308
Open
howlcipher wants to merge 2 commits into
Open
howlcipher wants to merge 2 commits into
howlcipher wants to merge 2 commits into
Conversation
This branch has not been deployed
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 join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Adds an internal process-local compact graph repair session so a model can propose existing bounded node replacements without retransmitting the full protected-node map in each direction. Trusted authority stays private; responses pass through the existing atomic graph/node/protected-region preconditions and full verification gate.
The initial AI-DX3 slice adds no CLI, source editing, durable signer/restoration, execution grants or production AST-path changes. Context digests identify retained authority; they are not signatures. Source mapping and corpus-wide repeated performance comparisons remain separate follow-ups, with no model-efficiency claim.
Validation: independent review and re-review; scoped focused and full HFIR tests; surrogate red/green regression; deterministic corpus behavior/payload measurements; repeated three-node allocation benchmark. Final local vet/build and uncached full tests passed (17.351s/15.704s/994.773s); final source/test/config hashes match. Public CI pending. The 53-node ambiguous localization grants no repair authority. Context+delta byte totals exclude common task/provider scaffolding and are not complete provider prompts.
Inherited hygiene bug72 remains a failed gate. The production count and duplicated lines match main (745 clones/12,826 lines); test count644 is unchanged, duplicated test lines10,372 vs10,366 disclosed. Owner acceptance is required before merging; PR307 approval does not carry over. No ceilings, exclusions or tests are weakened.