[napetrov]

Summary

before:  cell A  docker run --name dokimi-run-T-4242 ...
         cell B  docker rm -f dokimi-run-T-4242   <- removes A's container
after:   dokimi-run-T-4242-3fa9c1 / dokimi-run-T-4242-0b7e22

A suite runs every harness cell in one process, so dokimi-run-<task>-<pid> repeated whenever two cells ran the same task. Now the name ends in 3 random bytes (os.urandom(3).hex()). The dokimi-run- prefix stays, so cleanup globs still match.

Evidence

• New test test_two_runs_of_one_task_in_one_process_get_their_own_container: fails on main, passes here.
• tests/test_container_agent.py tests/test_task_formats.py: 234 passed, 1 skipped. ruff clean. docker_solver.py stays at its 927-line growth limit.

Merge Danger

Low. Anything that guessed the container name from the pid alone would no longer find it; nothing in the repo does.

🤖 Generated with Claude Code

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 92.02%. Comparing base (09be35b) to head (1459ab6).
❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #216   +/-   ##
=======================================
  Coverage   92.01%   92.02%           
=======================================
  Files         223      223           
  Lines       27768    27768           
  Branches     4563     4563           
=======================================
+ Hits        25552    25554    +2     
+ Misses       1570     1569    -1     
+ Partials      646      645    -1     

Files with missing lines	Coverage Δ
dokimi/dokimi/harnesses/docker_solver.py	97.71% <100.00%> (ø)

... and 1 file with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.