# Changelog

> **This is the PUBLIC form of this artifact. Findings are listed in full, but the details of SECURITY findings — which rule fired, in which file, on which line, and how to fix it — are deliberately withheld, and any secret-scanner results are excluded entirely. Where detail is absent here it was REMOVED FOR PUBLICATION; it is not missing from the analysis. The complete artifact is available from the repository owner.**

## Score

- CAI 54 → 54 (-0.0)
- Rubric changed (rubric-2026.10.5 → rubric-2026.11.2) — scores are not directly comparable.

## Lenses

- Code Health 75 → 75 (+0.0)
- Architecture 96 → 96 (+0.0)
- Maturity 55 → 54 (-1.0)
- Readiness 36 → 36 (+0.0)
- Security 83 → 86 (+3.2)
- Domain Modelling 100 → 100 (+0.0)

## Resolved (2)

- Inconsistent method naming and signature for bulk discard operations. `ClaimedExecution` and `Execution` both have `discard_all_in_batches` and `discard_all_from_jobs`, but the signatures differ slightly (`ClaimedExecution.discard_all_in_batches` takes no args implying it operates on the current context, while `Execution.discard_all_in_batches(batch_size)` takes an explicit batch size). This inconsistency in argument requirements for semantically identical operations across related types is confusing.
- Redundant/Confusing naming for recurring task registration. `SolidQueue.schedule_recurring_task` is a high-level convenience method that likely delegates to or is conceptually identical to `RecurringTask.create_dynamic_task`. Having two distinct entry points for the same logical operation (registering a recurring job) creates confusion about which API to use.

## New (1)

- Medium: security finding (details withheld)

## Changes since last survey

- 29 commits — 28 feature/other, 1 fixes

## By area

- app/models — 15 commits
- (root) — 7 commits
- lib/solid_queue — 3 commits
- test/models — 2 commits
- lib/solid_queue.rb — 1 commit
- test/integration — 1 commit

## Notable commits

- fix: Fix rescheduling recreated dynamic tasks
- change: Acquire concurrency locks by key when dispatching several jobs
- change: Bump solid_queue to 1.7.1
- change: Bump solid_queue to 1.7.2
- change: Correct which characters grow when the failed execution error is escaped
- change: Create executions from the job record instead of its id
- change: Don't eager load the CLI
- change: Escape LIKE metacharacters in queue prefixes
- change: Escape queue prefixes with a character no database treats specially
- change: Explain how async mode handles TERM and QUIT with jobs in flight
- change: Explain how to remove recurring tasks left behind by an empty configuration
- change: Fit failed execution errors into size-limited error columns
- change: Give concurrency slots back only to the semaphore they came from
- change: Ignore schema lookups when checking that enqueuing doesn't reload the job
- change: Keep batch completion errors from blocking subsequent jobs
- change: Let the job answer for the concurrency slot it holds
- change: Move the concurrent dispatching test next to the other semaphore lock tests
- change: Pin json below 3 for Rails 7.1 and 8.0
- change: Place the semaphore note under 1.7.2 in UPGRADING
- change: Quote the LIKE escape so MySQL accepts queue prefixes
- …and 9 more
