Merge pull request #6 from METR/feat/citing-scanner

feat: add citing_scanner emitting message and event references

Merge pull request #5 from METR/feat/hardcoded-429-simulation

Simulate 429s so adaptive concurrency can be tested

Merge pull request #3 from rasmusfaber/feat/network-sandbox-per-servi…

…ce-modes

feat(network_sandbox): support per-service network modes

Merge pull request #1 from METR/pip/add-failing-scanner

Add failing_scanner for deterministic scan-error testing