Merge feature/SB-0MNI23PTF005PGQ6-add-cli-runner-tests into main

Author: Sorra

.github/workflows/test-coverage.yml

@@ -0,0 +1,25 @@
+name: Test Coverage
+
+on:
+  push:
+    branches: [ main ]
+  pull_request:
+    branches: [ main ]
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Use Node.js 20
+        uses: actions/setup-node@v4
+        with:
+          node-version: 20
+
+      - name: Install dependencies
+        run: npm install
+
+      - name: Run tests with coverage
+        run: npm run test:coverage

README.md

@@ -100,6 +100,7 @@ See [Local Setup](#local-setup) below for detailed instructions.
 - `npm run start` - Run compiled bot from `dist/`
 - `npm run lint` - Type check without emitting files
 - `npm run test` - Run unit tests
+ - `npm run test:coverage` - Run tests with coverage and enforce >=80% coverage for the bot/ module
 - `npm run db:migrate` - Apply SQL migrations to PostgreSQL
 
 ## CLI Commands

docs/coverage-followups.md

@@ -0,0 +1,21 @@
+Coverage follow-ups for src/bot (SB-0MNFT3MI1005HXOD)
+
+Summary of local run
+
+- Test coverage for src/bot (cli-runner.ts) is currently 76.22% (statements/lines/functions/branches below 80%).
+- CI will now run coverage and fail if threshold is not met.
+
+Proposed follow-up work items
+
+1) Add unit tests to cover uncovered branches and lines in src/bot/cli-runner.ts
+   - Focus areas: error handling branches, timeout behavior, child process termination paths.
+   - Estimated effort: small (2-4 tests per area).
+
+2) Add integration tests for edge-case CLI outputs (invalid NDJSON lines, stderr-only errors)
+   - Use vi.mock for child_process.spawn to simulate CLI behaviour.
+   - Ensure last-event mapping and failed path exercised.
+
+3) Add test utility to reset and restore setCliPath / process.env between tests
+   - Ensure no cross-test leakage of CLI path.
+
+If you want, I can create these follow-up work items in the worklog and begin implementing the highest-priority tests.

docs/feature-requests/openbrain-playwright-fallback-retrieval.md

@@ -0,0 +1,47 @@
+# Feature Request: Playwright Fallback for Content Retrieval Failures
+
+## Summary
+
+OpenBrain should use a headless Playwright browser as a fallback when the standard retrieval path fails to return usable content.
+
+## Problem
+
+Some pages are rendered mostly or fully through JavaScript. The current retrieval path can fail on these pages, returning empty or low-value content, which then causes extraction and summarization to fail.
+
+## Why this matters
+
+- Shared links in Discord can fail ingestion even when the page is valid in a normal browser.
+- Community knowledge capture quality drops when JavaScript-heavy sites cannot be indexed.
+- Operators currently have no reliable fallback for blocked or dynamically rendered content.
+
+## Current architecture context
+
+SourceBase (Discord bot) does not implement page retrieval directly. It delegates ingestion to OpenBrain via `ob add`.
+
+Because of this separation, the fallback logic belongs in OpenBrain, not in SourceBase.
+
+## Requested behavior
+
+1. Attempt the current retrieval path first (fast path).
+2. If retrieval or extraction fails due to known failure modes (for example: blocked response, script-rendered content, empty extraction), retry with Playwright.
+3. Feed Playwright-derived HTML/content into the existing extraction flow so downstream behavior stays consistent.
+4. Emit clear telemetry/event data when fallback is used.
+
+## Acceptance criteria
+
+- OpenBrain continues to use the existing retrieval path by default.
+- OpenBrain invokes Playwright fallback only when the fast path fails or yields unusable content.
+- JavaScript-heavy pages that previously failed can now be ingested successfully when fallback is enabled.
+- Timeout and resource controls exist for browser-based fallback.
+- Logs or progress output make fallback activation visible for diagnostics.
+- Automated tests cover success and failure cases for both fast path and fallback path.
+
+## Non-goals
+
+- Implementing custom retrieval logic in the SourceBase Discord bot.
+- Making Playwright the primary path for all URLs.
+
+## Related work
+
+- Work item: Use playright retrieve content if existing retrieval path fails (SB-0MNHOYCUK000RALJ)
+- Historical draft (deleted): Add Playwright fallback for blocked content extraction (SB-0MN4FHM2B0PI69AE)