# Example workspace: click around Propel Source: https://runpropel.com/demo Example workspace # Click around a workspace. This is Propel working on Specbot, an example product that turns specs into test suites. Everything is example data, and nothing you click sends anything. Approve drafts in the queue, watch a job at work, read results with their ranges, and see how far each channel may go. Follow the five-step tour, or explore on your own. Example workspace · everything here is example data[Join the waitlist →](https://runpropel.com/waitlist) S Specbot 1 product New task Inbox 4 Library Results Automations Settings 1 of 5 · Tour Start here You talk to Propel like a colleague. The three doors show what it is doing for this product right now. Back Next 4 decisions waiting · 12 conversations started this week · cited in 7 of 20 buyer questions · read Monday · 6 posts out this week Get customers 12 conversations started this week Get found 7 of 20 buyer questions mention you Get talked about 2 of 6 posts beat your usual We ship GitHub Actions support on Tuesday. Get us ready. On it. I'll read what changed, check what AI assistants say about running spec tests in CI, write the launch thread and find QA leads who use GitHub Actions. That's about 180 credits. Watch the job → Ask Propel anything… Example: typing is off ## Needs you Launch thread for X · ranked 1 of 6 CHANGELOG.md "We deleted 4,000 snapshot tests. Then we wrote 300 from the spec…" Approve Edit Not this one 40 emails to QA leads specbot.dev/pricing "Your team writes tests after the bug turns up. What if the spec wrote them first?" Approve Edit Not this one Docs page: running spec tests in CI README.md The page ChatGPT, Claude and Perplexity are missing. Goes to your docs site. Approve Edit Not this one Priya replied: interested reply "Can you show me how it handles flaky tests?" Your booking link is ready. Approve Edit Not this one ## Get us ready for Tuesday's launch Working Quote: 180 credits · approved by you · 2 of 4 pieces ready Researcher Read 3 similar launches on Hacker News and Product Hunt Writer Thread ranked; email and blog post drafted Prospector 46 found through your Apollo key, 40 new Editor Checked every draft against your claims - ✓ Read what changed in the repo this week 4s - ✓ Asked ChatGPT, Claude and Perplexity how to run spec tests in CI 11s - ✓ Wrote 6 versions of the launch thread and ranked them 1m 08s - ✓ Checked every claim and number against its source 9s - … Waiting for your yes on 4 pieces now ## This week, for Specbot Get customers 12 conversations started · 40 emails, 5 replies, 3 interested Get talked about 2 of 6 posts beat your usual on X · measured at 48 hours Get found · how often AI answers mention you ChatGPT 31% Claude 22% Gemini 12% Perplexity 40% Grok 9% Bing 18% 120 answers per engine · read Monday · ranges show where the true figure probably sits ## Standing jobs Later Standing jobs are not in the public beta. They switch on after it. Every release Draft the launch: thread, email and docs page On Every Monday Check what AI assistants say about Specbot On When a competitor changes their pricing page Tell me what changed and draft a reply post On When search clicks swing 30% Find the page that moved and draft a fix On Every morning at 8 Send me a summary of what went out and what is waiting On ## How far each channel can go Limits read from the product's own rules. Steps are example progress. Search articles 1 2 3 4 Now: It posts unless you stop it · Limit: It runs on its own Pages for AI answers 1 2 3 4 Now: It posts unless you stop it · Limit: It runs on its own Brand posts 1 2 3 4 Now: It drafts, you approve · Limit: It runs on its own Your own opinion posts 1 2 3 4 Now: It drafts, you approve · Limit: It runs on its own Cold email 1 2 3 4 Now: It drafts, you approve · Limit: It posts unless you stop it Communities 1 2 3 4 Now: It suggests · Limit: It suggests Example product, people and numbers. The way it works, and the limits, are Propel's own.