predictleads-dashboard
Skill298 starsChanged 6 months agoMITView on GitHub
Use when a teammate wants to visually browse PredictLeads signals already cached in local SQLite — triggers include "dashboard for [domains]", "visualize signals for [list]", "show signals as a dashboard", "HTML view of [client lookalikes]", or any request to scan many companies' signals at a glance.
- Reads credentials
In this repository37 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- answer-linkedin-commentsSkill
- build-routineSkill
- campaign-dashboardSkill
- campaign-strategySkill
- claap-weekly-recapSkill
- cross-platform-schedulerSkill
- debuggerSkill
- enrich-with-signalsSkill
- find-lookalikesSkill
- fullenrich-content-engagersSkill
- fullenrich-event-attendeesSkill
- fullenrich-network-activationSkill
- fullenrich-plg-reverse-lookupSkill
- icp-importSkill
- import-leadsSkill
- integrate-with-yalcSkill
- launch-linkedin-campaignSkill
- lemlist-campaign-from-icpSkill
- list-adaptersSkill
- lost-deal-revival-agentSkill
- notion-reviewSkill
- personalize-messageSkill
- predictleads-lookalikesSkill
- predictleads-signalsSkill
- prospect-discovery-pipelineSkill
- provider-builderSkill
- qualify-leadsSkill
- research-prospectSkill
- run-competitive-intelSkill
- run-doctorSkill
- scrape-post-engagersSkill
- send-cold-emailSkill
- setupSkill
- show-routineSkill
- track-campaignsSkill
- yalc-orientationSkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 0 are distinctive, and 5 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 12
After running prospect-discovery-pipeline to scan all 10 finalists in one view
- Very specific: line 13
After bulk-enriching a campaign result set (signals:enrich --result-set) for a visual sanity check before outreach
- Very specific: line 29
Read the lead metadata into a Python dict (see existing template at ~/Desktop/predictleads-dashboard.html for shape).
- Very specific: line 38
Write to ~/Desktop/predictleads-dashboard-{client_or_topic}-{date}.html and open it.
- Very specific: line 44
If the script is missing, model the new one on the prior run captured at ~/Desktop/predictleads-dashboard.html (Apr 30 2026).
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
PredictLeads Dashboard (HTML viz)
Generates a single self-contained HTML page from cached signals in ~/.gtm-os/gtm-os.db. Cards per company with signal-count badges, top-signal callout, expandable detail (recent jobs, news, funding, tech stack, similar companies). Filter by vertical, sort by signal density or recency. Auto dark/light. Zero API calls.
When to use
- After running
prospect-discovery-pipelineto scan all 10 finalists in one view - After bulk-enriching a campaign result set (
signals:enrich --result-set) for a visual sanity check before outreach - Sharing signal context with a non-technical teammate (open the HTML, no CLI knowledge needed)
Don't use when: you only have signals for 1–2 companies (just use signals:show); signals haven't been pulled yet (run signals:fetch first).
How to invoke
The dashboard is built by a small Python script. Pass a list of domains and an optional list of pre-built lead cards (name + title + LinkedIn URL).
Inputs the skill needs
- List of domains (must already be in
company_signalstable) - Optional per-domain lead metadata:
{ company, vertical, geo, lead_name, lead_title, linkedin }
Build steps
- Read the lead metadata into a Python dict (see existing template at
~/Desktop/predictleads-dashboard.htmlfor shape). - Query SQLite for each domain:
SELECT signal_type, COUNT(*)for badge counts- Top 8 jobs by
event_date DESC - Top 8 news by
event_date DESC - Top 5 financing events
- Top 12 technologies
- Top 10 similar_companies sorted by
payload.score
- Render the HTML template (see
Implementationbelow) with embedded JSON. - Write to
~/Desktop/predictleads-dashboard-{client_or_topic}-{date}.htmland open it.
Implementation
A Python generator script lives at scripts/predictleads-dashboard.py (when committed). It reads from ~/.gtm-os/gtm-os.db, accepts a JSON config of leads, and emits a self-contained HTML file.
If the script is missing, model the new one on the prior run captured at ~/Desktop/predictleads-dashboard.html (Apr 30 2026). Key visual elements to keep:
- Per-company card with company name + vertical tag (color-coded) + domain
- Marketing lead pinned at top of each card with LinkedIn link
- 5 signal-type badges with counts (
jobs / funding / news / tech / similar) - "Top signal" callout with the most recent dated signal across types
- Expandable detail section (jobs/news/financing/tech/similar lists)
- Filter pills (All / vertical) + sort pills (density / recency / vertical)
Quick reference
# After signals:fetch has populated the cache for the domains you care about
python3 scripts/predictleads-dashboard.py \
--domains personio.com,oysterhr.com,...,mirakl.com \
--leads-json /tmp/leads.json \
--out ~/Desktop/predictleads-dashboard.html
open ~/Desktop/predictleads-dashboard.htmlCommon pitfalls
- Empty cards: signals haven't been fetched yet. Run
signals:fetch --domain Xfirst. - News headlines blank: PredictLeads news payloads use
summarynottitle. The template's display logic falls throughpayload.title || payload.headline || payload.summary. - Tech stack shows blanks: technology names live in JSON:API
relationships.technology.data.idresolved viaincluded[]. The normalizer inpredictleads-enrichment.tsalready promotespayload.technologyto a top-level string. Older signals fetched before the normalizer fix may have empty tech rows; re-fetch with--no-cache. - Similar companies show only score: same root cause — re-fetch with
--no-cacheto populate thesimilar_companyfield with the resolved domain.
Required env
None for generation (it's local-only). The signals must already be cached, which means PREDICTLEADS_API_KEY + PREDICTLEADS_API_TOKEN had to be set when the cache was populated.
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.

