v0.4.0 — Security + Language-aware + Ecosystem

Major polish release. See CHANGELOG.md for the full list.

Security

• Cache path traversal hardening
• 5MB HTTP response cap

Fixed — language-aware fallbacks (TR / ES / FR / DE)

• generate_outline synthetic H2s + must_cover are now per-language (no more English leakage in long non-EN outlines)
• Style wrappers (H2_STYLE_TEMPLATES, SUMMARY_H2, TOPIC_CONNECTOR) per language
• suggest_image_alts templates per language
• Tone jargon lists per language

Added

• Turkish suffix stemmer (stem_tr) for check_duplicates morphology awareness
• Pronoun families (sen vs siz, tu vs vous, du vs Sie, …)
• Stratified sitemap sampling
• Logging + input validation across all 14 tools
• examples/ folder with 5 realistic artifacts
• SECURITY.md, CODE_OF_CONDUCT.md, .github/dependabot.yml
• smithery.yaml + glama.json registry manifests, docs/REGISTRY_SUBMISSIONS.md
• Codecov integration
• Full 14-tool API reference in README + "Use as a Python library" + "Troubleshooting" + "What this MCP does NOT do"

Upgrade: uvx --refresh --from git+https://github.com/canberkys/seo-echo-mcp@v0.4.0 seo-echo-mcp

v0.3.0 — Image alts, urls param, cache, TR/DE passive

See CHANGELOG.md.

Added

• New tool suggest_image_alts (14 total). Flags missing/weak alts + proposes replacements from filename + keyword + nearest paragraph.
• analyze_site now accepts urls: list[str] — skip sitemap discovery for JS-rendered / blocked sites.
• Persistent cache for analyze_site at ~/.cache/seo-echo-mcp/<domain>.json (24h TTL default; SEO_ECHO_CACHE_DIR env var to override).
• Passive voice detection for Turkish + German in readability_report.
• CONTRIBUTING.md + issue/PR templates.

Upgrade: uvx --refresh --from git+https://github.com/canberkys/seo-echo-mcp@v0.3.0 seo-echo-mcp

v0.2.0 — Content creator expansion

See CHANGELOG.md for the full 13-tool list.