chore(metadata): adding badges, more tags, wider .npmignore

[allanbowe]

Issue

n/a

Intent

Improved packaging

Implementation

• more tags in package.json
• improved .npmignore (ignores mc_*)
• badges in readme

Checks

• Code is formatted correctly (sasjs lint).
• Any new functionality has been unit tested.
• All unit tests are passing (sasjs test).
• The PR desc or underlying commits follow the Conventional Commit standard

[4gl-reviewer]

Test Coverage Report

Ran npx @sasjs/cli c -t server on the PR branch.

Overall coverage: 120/253 macros (47%)

PR-changed macros

This PR is a metadata/packaging change (chore(metadata)). No .sas macro files were modified, so there are no PR-changed macros to assess for test coverage.

PR-changed file	Type	Coverage impact
.npmignore	packaging config	n/a — no macros
README.md	docs	n/a — no macros
package.json	packaging config	n/a — no macros
.agents/skills/sas/SKILL.md	docs	n/a — no macros
skills-lock.json	lockfile	n/a — no macros

No test coverage regression is possible from this PR since no runtime SAS code changed.

Generated by Hermes Agent (GitHub App)

[4gl-reviewer]

Hermes Agent Code Review

Verdict: Approve

This is a clean chore(metadata) PR — improved npm packaging, README badges, and a useful SAS skill doc note. No .sas macro files are touched, so there is no runtime/behavioural change and no test-coverage impact. Pre-review checks passed: mergeable: true, mergeable_state: clean, CI (Analyze) success, not a draft.

Critical

None.

Warnings

None.

Suggestions

• .npmignore:2 — mc_*.sas is correct (it targets the per-folder bundle files build.py emits: mc_base.sas, mc_ddl.sas, …), but since those files are not committed to git the pattern looks "dead" to anyone reading the ignore list in isolation. A trailing comment line (e.g. # build.py per-folder bundles) would prevent a future maintainer from deleting it. (Inline comment posted.)
• README.md — the removed [![GitHub closed issues]…] badge was the only badge linking into the issue tracker. Dropping it is a cosmetic call, but if issue-tracker visibility matters to contributors, consider keeping it (or the issues badge). Not blocking.

Looks Good

• package.json — removing "main": "index.js" is the right call: index.js has never existed in the repo (confirmed on main), so the field was pointing npm/Node consumers at a non-existent entry point. For a SAS-macro-only package there is no JS entry point, so omitting main is correct.
• package.json — the expanded keywords (sas-macros, sas9, sas-viya, datastep, etl, sasjs-cli, …) and the more descriptive description meaningfully improve npm search discoverability. The prefix hints (mf_*, mp_*, mm*) in the description are accurate (verified against the 408 .sas files: 88 mp_, 51 mf_, 45 mm).
• .npmignore — sasjsbuild/ and sasjsresults/ correctly exclude @sasjs/cli build output directories from the published tarball.
• .agents/skills/sas/SKILL.md — the new note about read-only automatic macro variables (&syscc, &syswarningtext, …) is accurate and a genuinely useful guardrail (trying %let syswarningtext=; does raise ERROR: Unable to assign value to a macro variable that is read only).
• skills-lock.json — the sas skill computedHash change is consistent with the SKILL.md content change above; the other skill hashes are untouched.

Test Coverage

Ran npx @sasjs/cli c -t server → 120/253 macros (47%). No macros changed in this PR, so no coverage delta. Full report posted as a separate comment.

Reviewed by Hermes Agent (GitHub App)