Skip to content

Commit ed47966

Browse files
TimDettmersclaude
andcommitted
Update coordinator guide: run only relevant tests, not full suite
Worker agents were running the full test suite (10+ min) which is wasteful when only a small area of code changed. Updated the completion workflow to instruct agents to run only relevant test files/functions. The full suite will be run separately later. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
1 parent 4ba0ccd commit ed47966

File tree

1 file changed

+16
-6
lines changed

1 file changed

+16
-6
lines changed

agents/coordinator_guide.md

Lines changed: 16 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -121,7 +121,7 @@ project before making changes so you can verify your setup works.
121121
- Information about existing PRs — what they change, whether they look correct
122122
- Anything else the worker agent should know
123123

124-
**5. Your recommended approach.** What you think the fix should look like. Be specific — name files, functions, line numbers. Frame it as guidance, not commands — the worker agent may find things you didn't and should use its own judgment.
124+
**5. Your recommended approach.** What you think the fix should look like. Be specific — name files, functions, line numbers. Frame it as guidance, not commands — the worker agent may find things you didn't and should use its own judgment. Include which specific test file(s) or test function(s) the agent should run to verify its fix — not the full suite.
125125

126126
**6. Completion workflow.** Every prompt file must include this section verbatim, with the issue number filled in:
127127

@@ -130,20 +130,29 @@ project before making changes so you can verify your setup works.
130130

131131
After implementing and verifying the fix:
132132

133-
1. **Commit** your changes with a message referencing the issue:
133+
1. **Run only the tests relevant to your change.** Do NOT run the full
134+
test suite — it takes 10+ minutes and will be run separately later.
135+
Instead, run the specific test file(s) that cover the code you changed:
136+
137+
pytest tests/test_autograd.py -v --tb=short -k "relevant_test_name"
138+
139+
If you wrote a new test, run that plus the existing tests in the same
140+
file to check for regressions in that area.
141+
142+
2. **Commit** your changes with a message referencing the issue:
134143

135144
git add <files>
136145
git commit -m "Fix <brief description> (#<NUMBER>)"
137146

138-
2. **Push** the branch:
147+
3. **Push** the branch:
139148

140149
git push -u origin fix/issue-<NUMBER>
141150

142-
3. **Create a pull request** with `gh pr create`. The PR body must
151+
4. **Create a pull request** with `gh pr create`. The PR body must
143152
include "Fixes #<NUMBER>" so GitHub auto-links and auto-closes the
144153
issue on merge. Describe what the fix does and how you verified it.
145154

146-
4. **Post to the bitsandbytes Slack channel** to notify the team.
155+
5. **Post to the bitsandbytes Slack channel** to notify the team.
147156
Write a temporary Python script to `/tmp/slack_notify.py` and run it:
148157

149158
import json, urllib.request, sys
@@ -245,7 +254,8 @@ whether it is correct and complete before implementing from scratch.
245254

246255
## When You Are Done
247256

248-
[the standard completion workflow section with issue number 1810 filled in]
257+
[the standard completion workflow section with issue number 1810 filled in.
258+
Remember: tell the agent to run only the relevant tests, not the full suite.]
249259

250260
## What NOT to Do
251261

0 commit comments

Comments
 (0)