Files
pipelines/examples/debug.md
T

138 lines
2.7 KiB
Markdown

# stage diagnose
```prompt
Investigate the reported problem in input.md without editing. Read the relevant
code and logs (search first, then read file ranges, not whole logs), reproduce the
failure with the narrowest safe command, and find the cause rather than the first
symptom. Record the evidence, the smallest credible fix, and anything you could
not verify. Return blocked when the problem cannot be reproduced or the needed
evidence is unavailable.
```
```tools
read
glob
grep
shell
```
```checks
none: read-only diagnosis; shell runs only the reproduction
```
```configs
provider=default
model=default
retries=2
timeout=1h
```
```edges
success -> fix
blocked -> #failure
error -> #failure
```
# stage fix
```prompt
Fix the diagnosed cause with the smallest coherent change. Follow the
repository's conventions, preserve unrelated work, and add or update a regression
test that would fail before the fix. Run that test and other focused checks for
the touched code; the committer runs the full checks once. Report the exact
commands and actual results. Do not commit, push or deploy.
```
```tools
read
glob
grep
edit
shell
```
```checks
none: focused checks only; the committer runs the project's required checks once
```
```configs
provider=default
model=default
retries=2
timeout=1h
```
```edges
success -> verify
blocked -> #failure
error -> #failure
```
# stage verify
```prompt
Independently verify the fix. Re-run the original reproduction and the new
regression test, and read the diff (`git diff --stat` first). Do not run the full
suite and do not edit. Return success only when the original failure is gone;
otherwise return rejected with concrete evidence.
```
```tools
read
glob
grep
shell
```
```checks
none: read-only verification; runs only the reproduction and regression test
```
```configs
provider=default
model=default
retries=2
timeout=1h
```
```edges
success -> committer
rejected -> #failure
error -> #failure
```
# stage committer
```prompt
Commit this run's change locally. Inspect `git status` and `git diff --stat`,
then the staged diff; stage only this run's work, never secrets or generated
files. Run the project's required checks once with quiet or summary output and
report the commands and results; return blocked with the failing output when one
fails. If nothing changed, say so rather than creating an empty commit. Do not
push, tag, deploy or rewrite history.
```
```tools
read
glob
grep
shell
```
```checks
none: generic starter has no project commands; with delivery=true list them here for the server to run
```
```configs
provider=default
model=default
retries=2
timeout=1h
```
```edges
success -> #success
blocked -> #failure
error -> #failure
```