Add event-follow-up skill — CSV → personalized post-event emails #82
Add event-follow-up skill — CSV → personalized post-event emails #82jay-sahnan wants to merge 1 commit intomainfrom
Conversation
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
There was a problem hiding this comment.
Cursor Bugbot has reviewed your changes and found 3 potential issues.
❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.
Reviewed by Cursor Bugbot for commit 13eec13. Configure here.
| if (depth === 'deeper') { | ||
| const r3 = bbSearch(`"${name}" github`); | ||
| const r4 = bbSearch(`"${name}" site:x.com OR site:twitter.com`); | ||
| out.links = { ...out.links, ...harvestLinks([...(r3.results || []), ...(r4.results || [])]) }; |
There was a problem hiding this comment.
Spread overwrites previously-discovered links with nulls
Medium Severity
The harvestLinks function always initializes all five link keys (linkedin, x, github, blog, podcast) to null. When lane 3–4 results are merged via { ...out.links, ...harvestLinks(newResults) }, every key from the new object overwrites the old one — so links discovered in lane 1 (e.g. a LinkedIn URL found by searching) that aren't re-found in lane 3–4 results get silently replaced with null. Only non-null values from the new call should overwrite existing entries.
Additional Locations (1)
Reviewed by Cursor Bugbot for commit 13eec13. Configure here.
| out.links = { ...out.links, ...harvestLinks([...(r3.results || []), ...(r4.results || [])]) }; | ||
| } | ||
|
|
||
| console.log(JSON.stringify(out, null, 2)); |
There was a problem hiding this comment.
Script enrich_person.mjs is unused dead code
Low Severity
enrich_person.mjs is not referenced by SKILL.md, workflow.md, or any other file in the skill. The person-enrichment workflow (Step 9) instructs subagents to run bb search directly via Bash — this script is never invoked. It appears to be leftover scaffolding that adds maintenance burden and contains its own link-merging bug.
Reviewed by Cursor Bugbot for commit 13eec13. Configure here.
|
|
||
| <footer> | ||
| <svg width="16" height="16" viewBox="0 0 100 100" fill="none" xmlns="http://www.w3.org/2000/svg"><rect width="100" height="100" rx="8" fill="#F03603"/><path d="M36 72.2222V27.7778H51.2381C57.5873 27.7778 62.6667 32.8571 62.6667 39.2063V41.746C62.6667 44.6667 61.5873 47.3968 59.7461 49.3651C62.2858 51.4603 63.9366 54.6349 63.9366 58.254V60.7936C63.9366 67.1428 58.8572 72.2222 52.508 72.2222H36ZM42.3493 65.873H52.508C55.3651 65.873 57.5873 63.6508 57.5873 60.7936V58.254C57.5873 55.3968 55.3651 53.1746 52.508 53.1746H42.3493V65.873ZM42.3493 46.8254H51.2381C54.0953 46.8254 56.3175 44.6032 56.3175 41.746V39.2063C56.3175 36.3492 54.0953 34.127 51.2381 34.127H42.3493V46.8254Z" fill="white"/></svg> | ||
| Generated by <a href="https://github.com/anthropics/skills">company-research</a> · Powered by <a href="https://browserbase.com">Browserbase</a> |
There was a problem hiding this comment.
Report template footer references wrong skill and repo
Medium Severity
The template footer says "Generated by company-research" and links to github.com/anthropics/skills. The renderShell function in compile_report.mjs never replaces the footer, so all three main report pages (index.html, people.html, companies.html) display the wrong skill name and wrong GitHub organization. The per-company detail pages correctly say "event-follow-up" linking to github.com/browserbase/skills.
Reviewed by Cursor Bugbot for commit 13eec13. Configure here.
1 participant
Summary
event-follow-upskill: takes a CSV of event attendees and produces a per-person follow-up email +sales-readiness flag (HOT / WARM / NURTURE / COLD), so AEs can stop sending generic "great meeting you"
templates that get ignored.
event-prospectingandcompany-research(ICPtriage → deep-research → enrichment), but pivots the output: enrichment + sales-readiness scoring + email
drafting all happen in one combined subagent pass per attendee.
.gitignorerule (skills/*/profiles/*.jsonwith!example.jsonexception) so personalICP profiles can't accidentally leak into commits across branches.
Note
Medium Risk
Adds a sizeable new skill with new Node scripts that parse CSVs, call the
bbCLI, and generate HTML/CSV outputs; while mostly isolated, failures could surface as incorrect lead filtering/scoring or broken report generation.Overview
Adds a new
event-follow-upskill that takes an attendee CSV, scores companies against a user ICP, enriches selected attendees, assigns a sales-readiness bucket (HOT/WARM/NURTURE/COLD), and drafts per-person follow-up emails.Introduces new Node tooling for the workflow:
parse_csv.mjs(header auto-detection + normalization),extract_page.mjs(bb-backed page extraction with browse fallback), and a dedicatedcompile_report.mjsthat rendersindex.html/people.html/companies.html, per-company detail pages, and a CRM-importableresults.csv.Updates repo ignore rules to prevent committing per-skill personal ICP profiles (only
profiles/example.jsonis tracked), and adds skill-specific docs/templates (workflow governance, research/email rubrics, report template, fixtures).Reviewed by Cursor Bugbot for commit 13eec13. Bugbot is set up for automated code reviews on this repo. Configure here.