← Pipelines

research-and-report

Run web searches for a user-supplied query, fetch page contents, synthesize findings, and post a summary to Slack.

Schedule

Manual only

Steps

5 steps

Created

2026-03-01 22:24:57

Run History

Run ID Status Started Finished Error
caf8d352-22aa-4859-a67c-e86a248a51a6 success 2026-03-01 22:55:42 2026-03-01 22:56:25
50b411f8-d5c2-4c4a-87ca-bd1e3b7ea0c2 success 2026-03-01 22:53:19 2026-03-01 22:54:07
180afb75-34bf-4955-970a-cfa8a5896a47 success 2026-03-01 22:37:15 2026-03-01 22:38:00
875f13e6-d081-4f7c-84be-044e6a3354a8 success 2026-03-01 22:28:24 2026-03-01 22:28:56
68177fa8-49e4-48c1-9ac8-33a4aa528a0c failed 2026-03-01 22:25:35 2026-03-01 22:25:58 step s4 (llm_format): llm_format: openai: openai chat json: error, status code: 429, status: 429 Too...

Update Pipeline

Pipeline Steps

s1 web_search
limit
8
query
{{.query}}
s2 web_search
limit
5
query
{{.query}} related topics
s3 fetch_content s1 s2
s4 llm_format s3
prompt
You are an assistant that consolidates web search results and article contents about the user query. The input events may include title, link, snippet, and content_readable. Produce a single plain-text report for '{{.query}}' with: 1) A short headline summary (1-2 sentences). 2) A numbered list of the most relevant sources: for each, include title, link, and a one-sentence synthesis using the snippet and content_readable. 3) A final 2-3 sentence overall takeaway and suggested next steps or follow-up searches. Keep output concise and informative.
s5 slack s4
text
{{.text}}

Pipeline JSON