chore: add missing keywords entries to two string/* package.json files

[Planeshifter]

Description

What is the purpose of this pull request?

This pull request:

• Adds five missing generic keywords entries across two @stdlib/string/* package.json files so their sibling-shared metadata cluster is complete.

Namespace summary

• Target: @stdlib/string (58 direct child packages; 56 leaf packages + 2 namespace aggregators base and tools, which are excluded from the semantic vote because they have no lib/main.js).
• Features analyzed: file tree presence, package.json top-level key set, keywords array membership, directories/scripts/stdlib key sets, README section presence and order, test/ / benchmark/ / examples/ file naming, manifest.json shape, plus semantic-shape grep across lib/main.js covering error-construction (format usage), JSDoc @example / @returns / @throws / @param presence, and copyright-year distribution.
• Features with clear majority (≥75%): test/test.js, examples/index.js, lib/index.js, README.md, package.json, docs/types/index.d.ts, docs/types/test.ts (all 100%); benchmark/benchmark.js, lib/main.js, docs/repl.txt (96.6% — missing only from the two aggregators); every package.json top-level key (author, bin, bugs, contributors, dependencies, description, devDependencies, directories, engines, homepage, keywords, license, main, name, os, repository, scripts, types, version) at 84.5–100%; README sections Usage (100%), Examples (100%), Notes (77.6%), CLI (84.5%); bin-field presence (84.5%, perfectly correlated with the CLI README section — not drift); keywords values utility/utils/string (100%), stdlib/stdstring/utilities/util/str (98.3%); semantic: 55/56 leaves use @stdlib/string/format for error construction (100% except format itself); 100% carry @example, @returns, @throws JSDoc tags; 0% use string concatenation or template literals in throw new *Error(...).
• Features without clear majority (excluded from drift analysis): README section ordering (plurality 29% — no ≥75% sequence); README See Also (70.7% — also generator-owned per pre-validation gate); lib/validate.js presence (5.2% — only 3 packages use a separate validation module); manifest.json presence (0%); binding.gyp/src/*.c/.npmignore presence (0%); scripts and stdlib package.json subkey sets (empty across the namespace); JSDoc @example count per file (bimodal, no ≥75% mode).

string/num2words

Fixes keywords[0] in lib/node_modules/@stdlib/string/num2words/package.json from missing to "stdlib", matching the shared metadata template used across the namespace. Prior to this, 57/58 (98.3%) @stdlib/string/* packages carried "stdlib" as their first keyword; num2words already had every other cluster keyword (stdstring, utilities, utility, utils, util, string, str) but this one. One-line diff, brings conformance to 58/58.

string/to-well-formed

Correction adds stdstring, utilities, util, and str to to-well-formed's keyword list — each keyword sits at 57/58 (98.3%) across @stdlib/string/*, and to-well-formed was the lone outlier despite already carrying the cluster's siblings (stdlib, string, utility, utils). No semantic basis for the omission; insertions land next to their existing cluster-mates, no reordering.

Validation

Three independent agents reviewed the candidate corrections:

• Semantic-review (opus) — classified both candidates as confirmed-drift. Both target packages are ordinary leaf string utilities; the missing keywords apply generically and the omissions are asymmetric relative to already-present cluster-mates. No mathematical/semantic shape difference justifies the deviations.
• Cross-reference (opus) — confirmed no test assertion, example output, README, docs/repl.txt, docs/types/index.d.ts, or docs/types/test.ts under either outlier inspects the keywords array; grep for the specific added keywords returns no matches inside the target packages; parent string/README.md references the outliers only as signature links, with no keyword contract; no keyword lint policy in .github/PULL_REQUEST_TEMPLATE.md.
• Structural-review (sonnet) — confirmed both corrections; the recommended num2words insertion position (stdlib at index 0) is unambiguous across all reference packages, and the four to-well-formed insertions are placed minimal-diff next to each keyword's cluster-mate without reordering the pre-existing sequence.

Excluded from this run

• README section ordering, Notes section content, See Also links — either below the 75% majority threshold, or generator-owned and covered by pre-validation gate 7a.
• docs/repl.txt presence gap in aggregator packages base / tools — generator-owned artifact per pre-validation gate 7a.
• benchmark/benchmark.js / lib/main.js presence gap in aggregator packages — inapplicable per pre-validation gate 7b (namespace aggregators legitimately lack a benchmarkable single function).
• bin-field / CLI README section absence in 9 packages — presence is perfectly correlated (packages either have a CLI or don't), reflecting an intentional design difference, not drift.
• The bin key in code-point-at/package.json is "from-code-point": "./bin/cli", which mismatches the package's documented CLI command name (code-point-at). This looks like a copy-paste bug but a fix would change the installed executable's name — a public behavior change — so it is flagged as needs-human and not touched here.

Related Issues

Does this pull request have any related issues?

No.

Questions

Any questions for reviewers of this pull request?

No.

Other

Any other information relevant to this pull request?

This PR is the output of an automated cross-package API drift detection routine over the @stdlib/string namespace. Namespace was picked uniformly at random from eligible (≥8 direct child packages) namespaces. Detailed per-feature majority pattern, conformance percentages, outlier list, and validation-agent verdicts are in the local drift report (not committed to the repo). PR is intentionally left in draft state pending maintainer audit.

Checklist

Please ensure the following tasks are completed before submitting this pull request.

• Read, understood, and followed the contributing guidelines.

AI Assistance

When authoring the changes proposed in this PR, did you use any kind of AI assistance?

• Yes
• No

If you answered "yes" above, how did you use AI assistance?

• Code generation (e.g., when writing an implementation or fixing a bug)
• Test/benchmark generation
• Documentation (including examples)
• Research and understanding

Disclosure

This PR was authored end-to-end by Claude Code running an automated cross-package drift-detection routine: namespace selection, structural feature extraction (Python over package.json, README, and file trees), semantic-shape extraction (targeted grep over lib/main.js covering error-construction, JSDoc tags, throws/returns), three-agent drift validation (opus semantic-review + opus cross-reference + sonnet structural-review), and patch application. Corrections are strictly additive metadata (keywords array entries) — no runtime, signature, or test behavior is touched. PR is left in draft for human audit before promotion.

---

@stdlib-js/reviewers

---

Generated by Claude Code

[stdlib-bot]

Coverage Report

Package	Statements	Branches	Functions	Lines
string/num2words	$\\color{red}513/521$ $\\color{green}+98.46\\%$	$\\color{red}82/85$ $\\color{green}+96.47\\%$	$\\color{green}6/6$ $\\color{green}+100.00\\%$	$\\color{red}513/521$ $\\color{green}+98.46\\%$
string/to-well-formed	$\\color{red}71/75$ $\\color{green}+94.67\\%$	$\\color{red}9/11$ $\\color{green}+81.82\\%$	$\\color{green}1/1$ $\\color{green}+100.00\\%$	$\\color{red}71/75$ $\\color{green}+94.67\\%$

The above coverage report was generated for the changes in this PR.