Reuse capability discovery metadata across fenced runs - #61
Merged
Merged
Conversation
added 3 commits
September 15, 2026 12:14
…ity revision # Conflicts: # docs/reference/workspace-packages.md
first-assist
marked this pull request as ready for review
September 15, 2026 02:39
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
A later fenced run previously lost its discovery candidates even when the capability catalogue was unchanged. An optional bounded process-local cache now lets the same principal and AI conversation describe a prior candidate without another model-driven discovery call.
Cached entries contain discovery metadata only. Describe rechecks current catalogue fingerprints, principal and policy before issuing a fresh run-bound handle; execution permissions and fences are unchanged. TTL is configurable from one second to seven days, with a bounded oldest-first search count. The default remains disabled. graphql-orm-ai is 0.98.1; no data or GraphQL migration is needed.
Validation: all 16 capability-delivery tests and 428 provider-enabled library tests passed (4 explicit live tests ignored). Added coverage for cross-run reuse, fresh execution fences, revocation, principal/conversation isolation, fingerprints, expiry, eviction and disabled defaults. Warnings-denied all-target Clippy and Rustdoc, PascalCase naming, compile-only PostgreSQL/MSSQL checks, documentation/release-policy checks and an explicit package SemVer comparison passed (196 compatibility checks).
This compatibility commit is based on 17d42f8 so consumers can adopt the cache without moving their existing ORM/auth dependency universe. Merge this PR with a merge commit to retain that tested compatibility revision. README, changelog, migration notes and generated workspace inventory are updated.