Repository Issue Activity (beta)

#95 fix(proxy): fatal crash due to concurrent map read and write in proxy pool

#28 SSRF — `/v1/scrape` fetches loopback, private-network and cloud-metadata URLs and returns the response body to the caller

#32 Jobs are stuck in `processing` forever when they hit `JOB_TIMEOUT` — the failure is written with the already-expired job context

#87 `ScrapeSync` never times out when store lookups keep failing — the `continue` on the error path skips the deadline check

#85 title: Domain config `requestTimeoutMs` is a no-op — `HandlerRequest.Timeout` is set but never read by any handler

#83 Real scrape errors are replaced by a browser-connection error — every failure reports `ECONNREFUSED :9222` instead of the actual HTTP status

#81 Batch job reports `"status": "completed"` when every URL in the batch failed

#72 Domain-config changes can remain unapplied for up to 60 seconds

#71 Add automated webapp tests for API client, polling, and recent dashboard flows

#70 Postgres batch reads silently discard child rows when scanning fails

#69 Batch creation returns success while silently dropping child jobs after a persistence error

#74 Health endpoint reports ok while the job store is unavailable

#73 Canceled job contexts still run fallback handlers and configured retries

#39 Dashboard "API Endpoints" card is missing /v1/telemetry/status and /health

#38 Web dashboard duplicates job-tracking logic instead of using the useJobs hook

#30 No authentication on any endpoint plus `AllowOrigins: "*"` — any web page can drive the scraper and rewrite domain configs

#36 Web dashboard has no way to set the sync scrape timeout (backend supports it via PR #12)

#66 Batch scrape submissions provide no inline feedback on the Scrape page

#26 Server panics on startup when `DATABASE_URL` is not set — the documented zero-config quick start is broken

#58 `worker.Pool.Submit` blocks the request goroutine when the queue is full, and panics if it races `Drain`

#57 `MemoryStore` never evicts jobs — the default zero-config mode grows unbounded until OOM

#56 Batch results are returned in random order with meaningless `index` values when running without `DATABASE_URL`

#55 Data race in the proxy pool — `Score` fields are read outside the mutex while `RecordSuccess`/`RecordFailure` mutate them

#54 Domain config `requestTimeoutMs`, `priority` and `isEnabled` are accepted, persisted and documented but never applied

#53 `TestScrapeSyncTimeout` re-implements the timeout logic instead of calling it — `internal/http/handlers` is at 0% coverage