Repository navigation
perf(server): per-thread settlement and PR checks no longer rebuild the whole thread list - #13691
Effect Service Conventions: 1 violation
apps/server/src/orchestration/ThreadPullRequestReactor.ts: readSweepSnapshot takes an Effect service instance as a parameter instead of acquiring it from the Effect environment. An inline review comment requests the fix.
Details
Warning
- Invalid effort level "max" in front matter. Fell back to "low". Valid values are: low, medium, high.
Note
Your check run agent prompt is: .macroscope/check-run-agents/effect-service-conventions.md
More information about how Check Run Agents work can be found in our Help Center.
Reviewed all 2 changed files (merge base moved)
🔍 Agent activity — 5 tool calls
💭 Inspecting service shape I need to take a closer look at the relevant source and the shape of the service to identify any possible violations. It seems the
readSweepSnapshotfunction is taking a service shape parameter, but the production helper isn't from the construction factory. I’m wondering if the rule factory is taking a service dependency, and whether the helper could be seen as an injection. There's a bit of ambiguity here that needs clarification.
1. browse_code — apps/server/src/orchestration/ThreadPullRequestReactor.ts
2. browse_code — apps/server/src/orchestration/ThreadSettlementReactor.ts
3. browse_code — apps/server/src/orchestration/Services/ProjectionSnapshotQuery.ts
💭 Clarifying service dependencies I'm thinking about a helper that functions as a flag for service shape, but the wording is confusing: it says to "Flag a factory that takes ... as a parameter when that value is a service dependency." A helper isn't the same as a factory, is it? The terms like "production service construction" and "yield" suggest there's some mixing going on. I should clarify this, avoid confusion, and make sure to comment precisely to fix the ambiguity in multiple files.
4. modify_pr — Submitted review with 1 inline comment
💭 Structuring conflicts clearly I think I need to ensure that the final structured conflicts are precise. It seems like things are only clear when there are no findings. I'm wondering if I should consider something neutral or if it’s better to stick with a clear failure. But then again, maybe I should just accept that “may fail” is a valid option. It’s a bit tricky to navigate! I want to get it right.
5. browse_code — apps/server/src/orchestration/Services/ProjectionSnapshotQuery.ts
Agent Credits: 8 credits
Token usage: 12 uncached input · 52,601 cache-read input · 20,195 cache-write input · 704 output
Agent Credits may also include non-token charges from external tools such as web research.