Repository navigation
Conversation
|
Important Review skippedToo many files! This PR contains 606 files, which is 506 over the limit of 100. To get a review, reduce the PR to 100 files or fewer by splitting it into smaller PRs or changing its base branch. Upgrade to a paid plan to raise the limit. Usage-priced reviews support at most 300 files. ⚙️ Run configuration
⛔ Files ignored due to path filters (4)
📒 Files selected for processing (606)
You can disable this status message by setting the
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Gavel summary
Totals: 0 passed · 0 failed · 0 skipped · - |
|
Review the following changes in direct dependencies. Learn more about Socket for GitHub.
|
Gavel summary
Totals: 6252 passed · 0 failed · 12 skipped · 6m41s |
Claude-Session-Id: c1750ebd-c2e1-43bc-8962-5499fa2e65d7
A decoded `"permissions": {}` marked /permissions explicitly present, and
the explicit-path pass then overwrote the merged section with the zero
struct — dropping every inherited value the override never named, so a
request that merely mentioned the section erased the project's tool policy.
An explicit path naming an all-zero struct now merges field-wise instead of
replacing. Every value the override actually supplied inside the section
still carries its own explicit path and still wins, including a
deliberately emptied collection like `"tools": {}`.
Claude-Session: https://claude.ai/code/session_01QemSU28zWbxmrRuNKfoFPb
Claude-Session-Id: 2d8cf8c1-0626-49fd-b02b-654e7c1554fe
…s only A spec layer is a source of defaults: the last layer naming a field owns it, and no layer restricts what a later one may select. A posture that must survive the whole stack is applied by its owner as the final layer rather than hung off an earlier layer as a ceiling. Deletes RuntimeConstraints and everything that folded it — the permission ceiling lattice, the restrictive model catalog, run limits and usage quotas, ValidateRuntimeConstraints, and the aichat/promptrun enforcement paths that read them. Nothing outside tests ever produced a model or quota constraint, and the permission lattice ranked only four of the six permission modes, so a stack pairing an unranked posture (auto, dontAsk) with any other refused to compose at all — one such pairing took down a whole run-context response. ComposeSpecLayers is now the fold alone: order the layers, merge the specs, record provenance, validate the effective budget. Claude-Session: https://claude.ai/code/session_01QemSU28zWbxmrRuNKfoFPb Claude-Session-Id: 2d8cf8c1-0626-49fd-b02b-654e7c1554fe
…nventory and Go DTOs Claude-Session-Id: 01a08ec3-abce-7663-b1a3-546c0b710945
Claude-Session-Id: 01a09053-85e4-7382-9b30-5a4048a9c95f
Claude-Session-Id: d72dec9c-ff9c-4f1e-b821-264063be5727
Claude-Session-Id: c7ef9a1a-c38c-4d0e-bf53-db8d1829d974
…nt mapper config Claude-Session-Id: 01a0a813-4df1-72f0-b045-3c31d672c8c0
…t-in defaults Claude-Session-Id: 01a0a813-4df1-72f0-b045-3c31d672c8c0
Claude-Session-Id: 01a0a813-4df1-72f0-b045-3c31d672c8c0
…n chat Claude-Session-Id: 01a0a813-4df1-72f0-b045-3c31d672c8c0
Claude-Session-Id: 01a0a813-4df1-72f0-b045-3c31d672c8c0
Claude-Session-Id: 01a0a813-4df1-72f0-b045-3c31d672c8c0
Claude-Session-Id: 01a0b3ab-29e0-7ee1-aa0e-89f7b48065a7
Claude-Session-Id: fe63c69c-4e6a-4230-86da-fca899327996
Claude-Session-Id: 01a0cc87-ee7d-7721-baca-11e38c8badba
Claude-Session-Id: c7d6497a-8a15-4f2f-b203-32fa7f360cac
Refactor Workspace to use a structured WorktreeState type instead of scalar Branch/Base fields. This provides clearer separation between isolated worktree runs and shared checkouts, and enables richer teardown reporting of worktree lifecycle (kept/removed, branch deleted, uncommitted changes). BREAKING CHANGE: Workspace.Branch and Workspace.Base removed; use Workspace.Worktree.Branch and Workspace.Worktree.Base instead. Workspace.Worktree is nil for runs in the caller's checkout.
The saved-defaults spec pinned claude-sonnet-5, which broke when the catalog promoted the sonnet alias to claude-sonnet-5-5. The alias target is pinned by the registry's own tests; this spec only checks the saved alias reaches resolution. Claude-Session-Id: f22d5af5-750c-4f54-a30f-57e8060a353b
An async Agent call returns at once, so the SDK settled the main turn while its subagent still ran. Captain ended the turn on that first result: the chat request closed the provider, the bridge shut down and killed the subagent, and persisting the turn failed on the subagent's unfinished tool call. - End a turn on the CLI's session_state_changed idle (opted into with CLAUDE_CODE_EMIT_SESSION_STATE_EVENTS), merging the SDK results it settled: usage sums, cost is the query's running total. - Carry parent_tool_use_id as Event.ParentToolCallID and AI SDK toolMetadata, exempt subagent calls from the turn's completeness checks, and close any the turn outlived with an explicit error. - Drop subagent narration from the parent's answer. - Dedupe streamed text by content: the SDK now sends one assistant message per completed block, so position-based dedupe printed every reply twice. - Let the bridge exit on shutdown before stopping it, so a clean run's task is not recorded as cancelled. Bumps the bridge protocol to 3. Claude-Session-Id: f22d5af5-750c-4f54-a30f-57e8060a353b
# Conflicts: # pkg/aichat/approval_execution.go # pkg/aichat/approval_execution_test.go # pkg/aichat/execution_database_authority.go # pkg/aichat/runtime_profile.go # pkg/aichat/runtime_profile_ginkgo_test.go # pkg/api/runtime_constraints.go # pkg/api/spec_layers.go # pkg/api/spec_layers_ginkgo_test.go # pkg/cli/webapp/src/App.tsx # pkg/cli/webapp/src/shellHelpers.ts # pkg/promptrun/preflight_constraints_ginkgo_test.go
Claude-Session-Id: 09629306-a4eb-413f-b217-1a5868fff340
Claude-Session-Id: ea298832-b5e2-4688-bcf2-785faf512651
…aming Enable schema validation retries for streaming responses by resuming provider sessions on repair turns. In addition, enable session reuse in agent verify loops to preserve context across attempts. Claude-Session-Id: b196d15e-247c-4cf2-947f-a50b32a85178
Update flanksource commons, commons-db, and clicky packages to latest versions. Adapt CLI live session unit test to use Codex provider telemetry.
Claude-Session-Id: 839416bd-5659-4f49-bc2e-eb0ed8c2b2a7
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.
No description provided.