[integrations] Chrome capture: skip per-turn LLM classification on bulk sync

[eazene]

What

Bulk sync captures from the Chrome capture extension now send skip_classification: true on the /ingest payload. Manual captures are unchanged and still classify.

Why

All three bulk-sync paths (Claude, ChatGPT, Gemini) funnel every turn through the gateway's smart-ingest pipeline. Without the flag, a 400-turn backfill fires ~400 per-item classification LLM calls on top of extraction — pure cost, since sync-mode turns arrive with full source provenance already attached (extension_platform, conversation IDs, page URL, client fingerprint).

Verified against a live smart-ingest deployment: A/B dry-runs of the same conversational turn with and without the flag produce identical extraction results (the flag only gates the later per-item classification pass), so no capture behavior changes other than skipping that call.

How

One line in the shared payload builder in processCaptureRequest (background/service-worker.js):

skip_classification: capture.captureMode === 'sync',

• All bulk paths set captureMode: 'sync'; manual clicks arrive as 'manual' → flag is false.
• The retry queue re-sends the stored payload whole, so the flag survives retries.
• Gateways that don't read the field simply ignore it (backward compatible).

Also bumps manifest.json to 0.6.1, adds a README changelog entry, and updates metadata.json (1.1.2 / 2026-07-14).

Testing

• node --check on service-worker.js; JSON validity on manifest/metadata.
• Live A/B smoke test against a deployed open-brain-rest → smart-ingest stack: identical extracted_count with/without the flag; provenance metadata persisted; dedup statuses unaffected.

🤖 Generated with Claude Code

https://claude.ai/code/session_01LxP7M99m4EF5Ve7rn3SNKH

[github-actions]

OB1 PR Gate

✅ Folder structure — All files are in allowed directories
✅ Required files — README.md and metadata.json found in all contribution folders
✅ Metadata valid — All metadata.json files passed JSON Schema validation
✅ No credentials — No API keys, tokens, or secrets detected
✅ SQL safety — No destructive SQL or core table modifications
✅ Category artifacts — Required file types present for each category
✅ PR format — Title follows [category] Description format
✅ No binary blobs — No oversized or binary files
✅ README completeness — All READMEs include Prerequisites, Steps, and Expected Outcome
✅ Contribution dependencies — All declared skill and primitive dependencies exist and are linked in README
✅ LLM clarity review — Covered by Claude PR Review workflow
✅ Remote MCP pattern — No local MCP server patterns detected — uses remote MCP correctly
✅ Tool audit link — Extensions/integrations link to the MCP Tool Audit guide
✅ Scope check — All changes are within the contribution folder(s)
✅ Internal links — All relative links in READMEs resolve to existing files

Result: All 15 checks passed! Ready for human review.

---

Post-Merge Tasks

These don't block merge — they're reminders for admins after this PR lands.

• Add Chrome Capture Extension to integrations/README.md
• Add Chrome Capture Extension to root README.md community contributions section
• Add Alan Shurafa (@alanshurafa) to CONTRIBUTORS.md
• Post in OB1 Discord #show-and-tell