Report lint warnings only on changed lines

[konard]

Fixes #103

Problem

Every Release run re-emitted the same local/no-changelog-comments warnings from files already on the default branch (33 in run 29434577797), hiding new actionable findings.

Solution

• scripts/lint-changed-lines.mjs: parses git diff -U0 into per-file changed line sets and filters ESLint results — errors always kept, warnings kept only when they land on a changed line (whole file for new/untracked files).
• scripts/lint.mjs: new npm run lint entry point that lints everything, reports filtered results, and prints how many warnings were suppressed. LINT_ALL_WARNINGS=1 (or npm run lint:all) restores the full output.
• eslint.config.js: disables the rule for docs/case-studies/** fixtures and for tests/no-changelog-comments.test.js, which intentionally contain historical wording.

Verification

tests/lint-changed-lines.test.js proves an unchanged baseline warning is not re-annotated while a warning on a newly changed line is, and that errors survive filtering.

$ npm run lint          # 0 problems (21 warnings on unchanged lines suppressed)
$ LINT_ALL_WARNINGS=1 npm run lint   # 21 warnings
$ node --test tests/lint-changed-lines.test.js   # 6/6 pass

[konard]

Working session summary

Done — PR #110 is ready: #110

• scripts/lint-changed-lines.mjs filters ESLint results: errors always reported, warnings only on lines changed vs. the base branch (whole file for new/untracked files).
• scripts/lint.mjs is the new npm run lint; it prints how many warnings were suppressed and honors LINT_ALL_WARNINGS=1 / npm run lint:all.
• eslint.config.js turns the rule off for docs/case-studies/** and the rule's own test.
• tests/lint-changed-lines.test.js (6 tests, passing) covers changed vs. unchanged findings and error preservation. Local lint/format checks pass.

---

This summary was automatically extracted from the AI working session output.

[konard]

🤖 Solution Draft Log

This log file contains the complete execution trace of the AI solution draft process.

💰 Cost: $0.892618

📊 Context and tokens usage:

• 42.9K / 1M (4%) input tokens, 8.5K / 128K (7%) output tokens

Total: (44 new + 29.3K cache writes + 770.9K cache reads) input tokens, 8.5K output tokens, $0.892619 cost

🤖 Models used:

• Tool: Anthropic Claude Code
• Requested: opus (claude-opus-4-8)
• Thinking level: off (disabled)
• Model: Claude Opus 4.8 (claude-opus-4-8)

📎 Log file uploaded as Gist (645KB)

• View complete solution draft log

---

Now working session is ended, feel free to review and add any feedback on the solution draft.

[konard]

🎉 Auto-merged

This pull request has been automatically merged by hive-mind.

• All CI checks have passed

---

Auto-merged by hive-mind with --auto-merge flag