~95% on SimpleQA (e.g. Qwen3.6-27B on a 3090). Supports all local and cloud LLMs (llama.cpp, Ollama, Google, ...). 10+ search engines - arXiv, PubMed, your private documents. Everything Local & Encrypted.
Python academiaanthropicarxivbravedeep-researchencryptionhome-automationhomeserverlocallocal-deep-researchlocal-llmmistralollamaopenaipubmedresearchresearch-toolretrieval-augmented-generationsearxngself-hosted
fix(ci): paginate sticky PR-comment lookups; make research comment sticky (#4333)
LearningCircuit112836790e21
diff --git a/.github/workflows/docker-tests.yml b/.github/workflows/docker-tests.ymlindex 8ffabd31fb..1ca509d96b 100644--- a/.github/workflows/docker-tests.yml+++ b/.github/workflows/docker-tests.yml@@ -397,11 +397,14 @@ jobs:</details>`;- // Find existing comment- const { data: comments } = await github.rest.issues.listComments({+ // Find existing comment. Paginate — on a long-lived PR the coverage+ // comment may be past the first page (listComments defaults to 30),+ // which would otherwise create a new comment every run.+ const comments = await github.paginate(github.rest.issues.listComments, {owner: context.repo.owner,repo: context.repo.repo,issue_number: context.issue.number,+ per_page: 100,});const botComment = comments.find(c =>diff --git a/.github/workflows/e2e-research-test.yml b/.github/workflows/e2e-research-test.ymlindex 8c265afd16..1ddad67cfa 100644--- a/.github/workflows/e2e-research-test.yml+++ b/.github/workflows/e2e-research-test.yml@@ -152,10 +152,48 @@ jobs:with:name: ${{ needs.research.outputs.comment-artifact-name }}- - name: Post comment+ - name: Post or update commentif: needs.research.outputs.success == 'true'- run: |- gh pr comment "$PR_NUMBER" --repo "${{ github.repository }}" -F comment.md+ continue-on-error: true # a GitHub API hiccup shouldn't fail the job+ uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0+ with:+ retries: 3+ script: |+ const fs = require('fs');+ const issue_number = Number(process.env.PR_NUMBER);+ // Key the marker by label so re-running the same label edits its+ // comment in place, while the other label keeps a separate comment.+ const marker = `<!-- ldr-research-results:${process.env.LABEL_NAME} -->`;+ const body = marker + '\n' + fs.readFileSync('comment.md', 'utf8');++ // Paginate — on a long-lived PR the existing comment may be past the+ // first page (listComments defaults to 30), which would otherwise+ // create a new comment every run instead of editing in place.+ const comments = await github.paginate(github.rest.issues.listComments, {+ owner: context.repo.owner,+ repo: context.repo.repo,+ issue_number,+ per_page: 100,+ });+ const existing = comments.find(+ (c) => c.user.type === 'Bot' && c.body.includes(marker)+ );++ if (existing) {+ await github.rest.issues.updateComment({+ owner: context.repo.owner,+ repo: context.repo.repo,+ comment_id: existing.id,+ body,+ });+ } else {+ await github.rest.issues.createComment({+ owner: context.repo.owner,+ repo: context.repo.repo,+ issue_number,+ body,+ });+ }- name: Remove label for re-triggeringif: always()