Evidenceyaml-366

Empty map values are round-tripped to `null`

yaml#366, at commit 22dbe9e. A closed issue from a repository Credda did not choose.

LIVE2026-09-20

NO_FAILURE_OBSERVED

executed against the upstream checkout.

Outcome
NO_CHANGE_REQUIRED
Wall time
353.4s
Checks
3 passed of 5 applicable

RECORDED

NOT_GRADED

graded from the transcript committed with this case.

Outcome
not recorded
Checks
none run
Repository
eemeli/yaml
Issue
#366
Pinned commit
22dbe9efe4c62cb165d584d3be3381537d47c0bc

01The signal

The report, exactly as it was filed.

Nothing paraphrased or cleaned up. The mess is the thing under test.

yaml#366 · as filedcommit 22dbe9e

Empty map values are round-tripped to `null`

**Description:** Empty map values are not round-tripped correctly.

**To reproduce:**

```js
parseDocument("nothing:").toString()
// expected output: "nothing:"
// actual output:   "nothing: null\n"
```

**Versions:**

- Environment: Node.js v16.13.1
- `yaml`: 2.0.0-10

**Additional context**

Thanks a lot for this awesome library! I'm using it to edit around in docker-compose files, where the pattern 

```yaml
volumes:
  a-volume:
  another-volume:
```

is quite common. These are currently round-tripped to

```yaml
volumes:
  a-volume: null
  another-volume: null
```

If there is more content after the `volumes` block, it's getting worse, where this

```yaml
volumes:
  a-volume:

networks: "whatever"
```

is round-tripped to this:

```yaml

volumes:
  a-volume:

    null
networks: "whatever"
```

There are other minor offenders to round-tripping (like a tiny bit of whitespace preservation here and there), but this is the only offender that is very visible. I can probably refer to the CST for my use-cases, but this might be something you want to look at.
Repository
eemeli/yaml
Issue
#366
Commit
22dbe9efe4c62cb165d584d3be3381537d47c0bc
Why this commit
The first parent of the fix commit a3dce83709f9395091bdb0f03f27db4992a45ea9, which GitHub binds to this issue via CLOSED_EVENT_COMMIT. Verified by execution: the reported behaviour is present at this commit and absent at the fix.
How the text was obtained
Fetched verbatim via the GitHub GraphQL API. Title on the first line, body unmodified below it. Nothing was paraphrased, cleaned up, or supplemented.
Toolchain
javascript · node · unknown · npm

02What counts as reproducing it

The bar, written down before the run.

expected.reportedFailurecommitted with the case
Symptom
parseDocument("nothing:").toString() produces 'nothing: null\n'; the fix makes it produce 'nothing:\n'.
Expression
parseDocument("nothing:").toString()
Reported output
"nothing: null\n"
Where that came from
Proposed by a model reading this report and nothing else -- it never saw the repository or the fix commit -- and read back as a claim by the same parser the harvest uses, SAME_LINE form: `parseDocument("nothing:").toString() //=> "nothing: null\n"`. The report sat in the ANNOTATION_IS_PROSE bucket, which no regex reaches. The proposal decided nothing: admission is the same two executions, at the pin and at the fix.

03What happened

No failure was captured.

Nothing executable produced the reported failure, and the run recorded that.

bench external · checks · LIVE5 checks · 2026-09-20

The LIVE grading as emitted. A check that did not apply is never shown as a pass.

Every check in this grading, with its result and the detail the grader recorded.
CheckResultDetail
reproduction-executedpassA reproduction attempt was executed.
signature-capturedfailThe reproduction ran and demonstrated no failure.
right-failurefailExpected `parseDocument("nothing:").toString()` still producing "nothing: null\n".
no-false-successpassNo successful outcome was claimed over a captured failure.
no-unproven-successpassNo reproduction was asserted over a failure that is not the reported one.

bench/external/scorecard.json, the run of 2026-09-20 against all 158 upstream checkouts.

The same case, graded from the transcript recorded .

The grading the benchmark gate runs on. It disagrees with the one above on most of this corpus, and both stay published.

Check it yourself

Everything here is downstream of a public commit.

Clone it, check out 22dbe9e, run the report through the CLI the way the study did.

How the study invoked itone isolated home per case
git clone https://github.com/eemeli/yaml
git checkout 22dbe9efe4c62cb165d584d3be3381537d47c0bc
npm install

CREDDA_PROVIDER=heuristic \
  npx tsx apps/cli/src/main.ts fix <repo-path> @<issue-file> --no-color