Every Monday I spend ~3 hours pulling competitor pricing and feature changes into a digest for the team.
System schema
Trigger
Each firing replays the 🎯 outcome (right) into a fresh ▶️ session in the sandbox — nobody at a keyboard.
LAUNCH.md fires the same agent with the same outcome.Worker
Researches the 5 named competitors' public sites, pricing pages, and changelogs; writes a weekly digest to /mnt/session/outputs/digest.md.
tools[] · agent_toolset_20260401
web_searchweb_fetchbashreadwriteeditglobgrep
resources[] · memory_store read_write competitor-history — last week's snapshot per competitor, so runs report changes, not re-descriptions.
mcp_servers[] + vault_ids[] — empty in v0, public web only. Notion + Slack sit in the version rail below.
skills[] — empty; plain markdown output.
Output & grading
digest.md — the weekly competitor digest, fetched from the run's outputs.
Your definition of done — rubric.content, graded by the platform every run.
| Run | Session | Rubric | Verdict | Note |
|---|---|---|---|---|
| — | — | — | no runs yet | The first launch fills this in. |
| Case | Known-good answer | Last verdict |
|---|---|---|
| week-of-may-25 | Lamis's hand-written digest | not run yet |
| week-of-jun-01 | Hand-written digest | not run yet |
| week-of-jun-08 | Hand-written digest | not run yet |
Case 1 is the v0 input; the rest are held back and re-run before promoting any new agent version.