docs: fix docstring parameter names in three components

[iridescentWen]

Related Issues

None — no prior issue. This is a docstring-only correction found while reading the code; happy to open one if you'd prefer that order.

Proposed Changes:

Three docstrings disagreed with their own signatures. I found these by walking every function in haystack/ with ast and diffing the :param names against the actual parameters.

1. LinkContentFetcher.__init__ — request_headers was undocumented (haystack/components/fetchers/link_content.py)

The signature takes request_headers: dict[str, str] | None = None and it is a real, functional parameter — _get_headers() merges self.request_headers into the outgoing headers. Every other init parameter is documented, so this one being absent reads as an omission rather than a deliberate choice. Added an entry describing it, including its precedence, which I took from the existing comment in _get_headers:

# client defaults -> component defaults -> user-provided -> rotating UA
base = dict(self._client.headers) if self._client is not None else {}
return _merge_headers(
    base, REQUEST_HEADERS, self.request_headers, {"User-Agent": self.user_agents[...]}
)

2. Sum.run — wrong parameter name (haystack/testing/sample_components/sum.py)

Documents :param value:; the parameter is values: Variadic[int]. One-character fix.

3. Threshold.run — missing parameter (haystack/testing/sample_components/threshold.py)

Takes value and threshold, documented only threshold. Added value.

3 files, +4/-1. Docstrings only — no executable line changed.

How did you test it?

No new tests: nothing executable changed, so there is no behavior to assert on. I ran the existing suites instead.

$ hatch run test:unit
5794 passed, 8 skipped, 322 deselected, 6 warnings in 33.98s

$ hatch run test:types
Success: no issues found in 382 source files

$ hatch run fmt
All checks passed!
273 files left unchanged

I also re-ran the ast walk after the change: the three functions now have no documented-but-nonexistent params and no undocumented ones.

Notes for the reviewer

On the release note: I did not add one. Per CONTRIBUTING, "Pull requests with changes limited to tests, code comments or docstrings ... can be labeled with ignore-for-release-notes by a maintainer in order to bypass the CI check." This PR is entirely docstrings, so I believe it qualifies — but only a maintainer can apply that label, so the release-note CI check will fail until someone does. Say the word if you'd rather I add a reno note instead and I'll push one.

Four near-misses I deliberately left alone. My scan also flagged these, and I checked each against the source and concluded they are correct as written — noting them so you know they were considered, not missed:

• ContextRelevanceEvaluator.run / run_async and FaithfulnessEvaluator.run / run_async document questions/contexts/predicted_answers while the signature is **inputs. Those names come from self.inputs in __init__, so documenting the real keyword names is more useful to a caller than documenting **inputs.
• LLM.run / run_async document messages, which lives in **kwargs. The code says why (# messages is intentionally omitted from the signature so the framework can treat it as required or optional depending on init configuration), and the docstring already says "Passed via **kwargs."

Three typos I did not touch. codespell finds trough→through, Docmentation→Documentation, and alogrithm→algorithm in releasenotes/notes/*.yaml. Those are notes for already-shipped releases, so I left them as historical record rather than rewriting them here. Let me know if you'd like a separate PR.

Checklist

• I have read the contributors guidelines and the code of conduct.
• I have updated the related issue with new insights and changes. — N/A, no related issue.
• I have added unit tests and updated the docstrings. — docstrings updated; no tests, see above.
• I've used one of the conventional commit types for my PR title (docs:).
• I have documented my code.
• I have added a release note file — requesting ignore-for-release-notes instead, see Notes for the reviewer.
• I have run pre-commit hooks and fixed any issue. — hatch run fmt clean, no files changed.

---

I used an AI assistant (Claude Code) to help write and verify this change. I have reviewed every line and run the tests above myself.

[vercel]

@iridescentWen is attempting to deploy a commit to the deepset Team on Vercel.

A member of the Team first needs to authorize it.

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[HaystackBot]

Hi @iridescentWen, thanks a lot for your contribution! 🙏

We noticed that the Contributor License Agreement (CLA) check (license/cla) hasn't passed yet, so we've temporarily moved this PR to draft and paused the review assignment.

To get your PR reviewed, please sign the CLA via the link in the license/cla check below (or in the CLA bot comment). As soon as the check turns green, this PR will automatically be marked ready for review again and a reviewer will be re-assigned.