research-lead
Skill0 starsChanged 46 days agoMITView on GitHub
Turn a lead (LinkedIn URL, name + company, or pasted profile) into a research package with relevant personalization and an outreach sequence, reviewed by a human before anything is sent. Run with /research-lead or ask to research a lead.
- Reads credentials
3 other projects use a version of this.
Adopted by
Most recently taken up by
- Celorus/celorus-plugintheir own version, first seen Nov 15, 2026
- jiahao-shao1/sjh-skillstheir own version, first seen Nov 8, 2026
- CatchIntent/skillstheir own version, first seen Jun 30, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- jiahao-shao1/sjh-skills★ 43same name
- CatchIntent/skills★ 0same name
- Celorus/celorus-plugin★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository24 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- analystSkill
- build-appSkill
- build-websiteSkill
- business-setupSkill
- content-pipelineSkill
- content-writerSkill
- email-assistantSkill
- email-digestSkill
- ksuite-setupSkill
- meeting-prepSkill
- memorySkill
- odooSkill
- plugin-builderSkill
- qfieldcloudSkill
- qgisSkill
- quarto-slidesSkill
- researchSkill
- r-sessionSkill
- schedulerSkill
- scrum-masterSkill
- skill-creatorSkill
- task-managerSkill
- weekly-reviewSkill
The file
We split this skill into the rules it gives an agent
4 rules. 0 are common, 3 are distinctive, and 1 is very specific to this project.
0
Common
Most of the 3 edited versions we read keep these.
3
Distinctive
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 26
Test for every fact: "Does this relate to a problem they're facing that we can solve?" If no, discard it.
- Very specific: line 41
Company research: follow assets/prompts/perplexity_research.txt with WebSearch / WebFetch (or mcp__searxng__* if configured).
- Distinctive: line 79
Never send a DM or an email yourself.
- Distinctive: line 90
No allowed hook: do not invent one.
“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Lead Research & Personalization (AIGORA)
What changed from the original template
The original pipeline chained five paid services (Relevance AI scraping through a
workflow owned by the template author, Perplexity, OpenAI, Airtable, Google Sheets,
Slack). In AIGORA the reasoning is done by you inside Claude Code, research uses
WebSearch/WebFetch (or a self-hosted SearXNG MCP), storage is a local CSV, and the
review card goes to kChat. Zero extra API key. The expert prompts were kept in
assets/prompts/ and are your instructions for each analysis.
Objective
Produce a research package with relevant personalization only. Relevant = relates to a problem they're likely facing that we can solve. Theater = personal but irrelevant (marathons, shared schools, hobbies). Test for every fact: "Does this relate to a problem they're facing that we can solve?" If no, discard it.
Inputs
- A LinkedIn URL, or name + company, or profile text pasted by the user.
context/my-business.md(replace{COMPANY_NAME}and the offer in the prompts with it).
LinkedIn pages usually block automated reading. Do not scrape LinkedIn. Use public sources (company site, press, public posts found by WebSearch) and ask the user to paste the profile text (About, Experience, recent posts) when needed.
Steps
- Profile: build
profile_data={full_name, first_name, company, headline, experiences: [{title, company, start, end}], recent_posts: [...], source_notes}. - Company research: follow
assets/prompts/perplexity_research.txtwith WebSearch / WebFetch (ormcp__searxng__*if configured). Cite URLs. Outputperplexity_data. - Analyses (you, in order), each following its prompt file and the schemas in
references/output-structures.md:lead_profile(assets/prompts/lead_profile.txt)pain_gain_operational(assets/prompts/pain_gain_operational.txt)dm_sequence(assets/prompts/dm_sequence.txt), max 300 characters per DMdm_quality_review: score 1-10 against the hard constraints of the DM prompt,approval_recommendationAPPROVE / REVISE / REJECT. Revise once if below 7.
- Assemble
.tmp/leads/<slug>/combined.json:{ "linkedin_url": "...", "profile_data": {...}, "perplexity_data": {...}, "lead_profile": {"data": {...}}, "pain_gain_operational": {"data": {...}}, "dm_sequence": {"data": {...}}, "dm_quality_review": {"data": {"overall_quality_score": 8, "approval_recommendation": "APPROVE"}}, "quality_flags": [] } - Review report (HTML, for the human):
python3 .claude/skills/research-lead/scripts/generate_review_report.py --data .tmp/leads/<slug>/combined.json - Store (local CSV
data/leads.csv, opens in R, LibreOffice or kSuite):
Then, if Odoo is configured (python3 .claude/skills/research-lead/scripts/save_lead.py --data .tmp/leads/<slug>/combined.jsonODOO_URLin.env) and the user approved the lead, offer to create the opportunity (preview first, then--confirm):python3 .claude/skills/odoo/scripts/odoo.py create-lead --from-json .tmp/leads/<slug>/combined.json [--stage ...]See theodooskill. - kChat review card (only if the user confirms): write a short Markdown summary
(who, why now, chosen hook, DM1, score) and post it with
python3 .claude/skills/ksuite-setup/scripts/kchat_post.py --file ....
Never send a DM or an email yourself. The human copies the approved message.
Batch
For a list of leads (CSV with linkedin_url or name,company), process them one by
one, skip those already in data/leads.csv, and give a summary table at the end.
Edge cases
- Private or unreachable profile: work from what the user pastes; note the limits.
- Company not found: say so; use headline and posts only; lower confidence.
- No allowed hook: do not invent one; recommend not contacting yet.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

