Add humanizer-ru skill (Russian-language AI-text editor)

[Vladimir-Human]

What

Adds humanizer-ru — a Russian-language editing skill that finds and removes traces of machine generation from Russian text and, on explicit request, rewrites it as a native speaker would.

This fills the empty Russian slot in the catalogue: finnish-humanizer is currently the only language humanizer here.

What the skill does

• 38 editing patterns (25 universal + 13 Russian-specific extensions) and 38 regex copy-paste markers (classes A/B).
• Every marker has an evidence registry entry: immutable source URL, access date, verbatim sample, evidence class (36 of 38 have full records; the remaining two are honestly closed as fixtures-only).
• Deterministic layer: 33 CI gates; blind pairwise evaluation (marker removal, fact preservation, false edits on a human corpus).

Declaration (important for review)

This is not a detector-bypass tool. The project explicitly declares "не средство обхода детекторов" (not a detector-bypass means) and states that false positives on living text are worse than missed machine text. The README includes a comparison table against same-named repos that openly advertise detector-bypass positioning.

Known automated-audit caveat: the skills.sh catalogue shows a red Snyk badge (E005) — a documented false positive on the Perplexity ppl-ai-file-upload marker (analysis pinned in https://github.com/Vladimir-Human/humanizer-ru/issues/46); the other two auditors pass.

Notes

• Bundle: SKILL.md + 16 reference files (~80 KB total), mirroring the structure of finnish-humanizer.
• Frontmatter validated with the project's own spec checker (0 warnings, strict mode). npm run skill:validate was not run locally — no npm in the contributor's environment; happy to fix anything your CI flags.
• Upstream repo: https://github.com/Vladimir-Human/humanizer-ru (MIT), released as v3.10.0.

[github-actions]

🟡 Contributor Reputation Check: MEDIUM risk

Check	Risk
Profile	MEDIUM
Credential audit	NONE

Maintainers: please review this contributor before merging.
See the workflow run for full details.
Automated check powered by AGT.

[github-actions]

🔒 PR Risk Scan Results

Scanned 19 changed file(s).

Severity	Count
🔴 High	0
🟠 Medium	0
ℹ️ Info	0

✅ No matching risk patterns were detected in changed files.

This is an automated soft-gate report. Findings indicate review targets and do not block merge by themselves.

[github-actions]

🔍 Vally Lint Results

✅ All checks passed

Scope	Checked
Skills	1
Agents	0
Total	1

Severity	Count
❌ Errors	0
⚠️ Warnings	0
ℹ️ Advisories	0

Summary

Level	Finding
ℹ️	✅ humanizer-ru (2/2 checks passed)
ℹ️	✓ [spec-compliance] All 1 skill(s) are spec-compliant.
ℹ️	✓ spec-compliance: All spec checks passed.
ℹ️	✓ [valid-refs] All file references across 1 skill(s) are valid.
ℹ️	✓ valid-refs: All file references resolve to existing files within the skill directory.
ℹ️	1 skill(s) linted, 1 passed

Full linter output

### Linting skills/humanizer-ru
✅ humanizer-ru (2/2 checks passed)
    ✓ [spec-compliance] All 1 skill(s) are spec-compliant.
        ✓ spec-compliance: All spec checks passed.
    ✓ [valid-refs] All file references across 1 skill(s) are valid.
        ✓ valid-refs: All file references resolve to existing files within the skill directory.

1 skill(s) linted, 1 passed

[aaronpowell]

Sorry, we're only currently accepting English-language skills at the moment.