gtm-tools-template
CLAUDE.md1 starsChanged 5 months agoMITView on GitHub
This repo is the operating manual for the [COMPANY] GTM motion. It's not application code — it's strategy docs, lead data, Claude skills, and scripts. Treat it like a wiki + a lightweight automation layer. All 22 skills live in .agents/skills/ — the cross-agent Agent Skills standard location read natively by Codex, Gemini CLI, Cursor, and others. .claude/skills/ contains per-skill symlinks to the same directories, so Claude Code auto-discovers them too. Edit skills only in .agents/skills/ (the symlinks make the…
In this repository23 other files it gives its agents.
- AGENTS.mdAGENTS.md
- article-writingSkill
- brand-voiceSkill
- deep-researchSkill
- email-personalizationSkill
- exa-searchSkill
- gtm-pipeline:company-enrichmentSkill
- gtm-pipeline:company-searchSkill
- gtm-pipeline:contact-filterSkill
- gtm-pipeline:demoSkill
- gtm-pipeline:outreachSkill
- gtm-pipeline:people-enrichmentSkill
- gtm-pipeline:people-searchSkill
- gtm-pipeline:pipelineSkill
- gtm-pipeline:setupSkill
- gtm-pipeline:signal-searchSkill
- investor-materialsSkill
- investor-outreachSkill
- lead-intelligenceSkill
- marketing-campaignSkill
- market-researchSkill
- research-opsSkill
- social-graph-rankerSkill
The file
We split this CLAUDE.md into the rules it gives an agent
4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 41
Don't add LinkedIn outreach automation.
- Distinctive: line 42
Don't optimize for sending volume.
- Distinctive: line 43
Don't add complex CRM integrations or pipeline tools yet.
- Distinctive: line 44
Don't draft op-eds in the voice of a SaaS founder.
≈ “Common” is an estimate: at least 1 in 5 of the 28,029 CLAUDE.md 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
CLAUDE.md — session guidance for gtm-tools
This repo is the operating manual for the [COMPANY] GTM motion. It's not application code — it's strategy docs, lead data, Claude skills, and scripts. Treat it like a wiki + a lightweight automation layer.
Project context (load-bearing)
- Product: [PRODUCT] ([DOMAIN]) — [PRODUCT_ONELINER]. AI built on one unified data model. Multiple pillars/modules across the operation: the entry pillar (live), adjacent intelligence pillars (live/beta), and roadmap pillars, unified by a shared analytics layer. NOT a single-feature tool with bolt-ons — a platform that enters through one budgeted wedge.
<!-- TODO: HQ location -->Demo CTA goes to "[DEMO_OWNER]" via [CALENDLY]. - Canonical traction: [METRIC_DEPLOYMENTS] deployed, [METRIC_USERS], [METRIC_LOI]. Anchor customer [ANCHOR_CUSTOMER] — we replaced the incumbent ([INCUMBENT]) entirely. [METRIC_COMPLETION] completion rate. (Record any phrasing changes — see
decision-log/YYYY-MM-DD-example-decision.md.) - Wedge (say this with investors, never in customer copy): enter through the budgeted, urgent, measurable entry pillar — close to the end user. The same trusted data layer then expands into the adjacent intelligence pillars. Crowded categories stay an expansion pillar, never the first-deal conversation.
- Investor framing: there is no $1B+ pure-play [INDUSTRY] operations software company; the fragmentation is the opportunity. See
investor/landscape-and-fundraise.mdfor comps, hardened stats, fundraise plan (~[METRIC_RAISE], [METRIC_ARR_MILESTONE] re-pricing milestone, accelerator timing, targets [INVESTOR]). - ICP priority (same first buyers; the difference is land-and-expand across pillars):
- (A) [ICP_A].
- (B) [ICP_B].
- (C) [ICP_C].
- Geography priority: [GEO_PRIMARY] first, then [GEO_SECONDARY] by referral.
- Buyer behavior: Heavy email users, light LinkedIn. Some segments use messaging apps heavily. Cold email is acceptable; warm intros via [ASSOCIATION] / design-partner referrals dramatically outperform.
- Volume: Aim for ~30–50 deliberate, personalized outreaches/week. Not 1,000/week. Most online GTM advice is calibrated for high-volume motions and will mislead at this stage.
- Team: Solo founder-led GTM, no prior GTM experience. [DEMO_OWNER] handles demos.
- Budget: ≤$500–700/mo on tools. Attend conferences but no booth sponsorships.
When making changes here
- This is documentation, not production code. Optimize for being readable by a non-engineer founder, not for elegance.
- Decisions get a decision-log entry. Any time you change the ICP, sequence, scoring weights, or tool stack, drop a file in
decision-log/YYYY-MM-DD-<topic>.mdexplaining what changed, why, and what we expect to see different. - Keep leads_raw.csv as-is. It's the Apollo export, our funnel input. Filtered/scored/enriched outputs live in separate files.
- Don't commit personal contact emails.
pipeline/leads/enriched.csvand*_with_emails.csvare .gitignored. The raw company list is fine. - Sender domain: outreach goes from
[OUTREACH_DOMAIN], NOT from the main[DOMAIN]. Protect the main domain's deliverability.
How to use the Claude skills in this repo
All 22 skills live in .agents/skills/ — the cross-agent Agent Skills standard location read natively by Codex, Gemini CLI, Cursor, and others. .claude/skills/ contains per-skill symlinks to the same directories, so Claude Code auto-discovers them too. Edit skills only in .agents/skills/ (the symlinks make the changes visible everywhere). Three groups:
email-personalization— the custom one we own. Takes a lead-list row and produces a personalized first-touch draft.gtm-*(10 skills) — pipeline mechanics: company/people/signal search, enrichment, contact filtering, outreach, demos. From keinsaasforever/gtm-pipeline-skills. Run/gtm-setuponce to configure API keys.- Research & GTM strategy (11 skills,
origin: ECCin frontmatter) — lead-intelligence, investor-outreach, investor-materials, market-research, deep-research, research-ops, exa-search, article-writing, brand-voice, marketing-campaign, social-graph-ranker. From affaan-m/ECC.
Both imported sets are MIT — see THIRD_PARTY_LICENSES.md.
What NOT to do in this repo
- Don't add LinkedIn outreach automation. Your buyers are not on LinkedIn the way SaaS CEOs are. Time spent there is mostly lost.
- Don't optimize for sending volume. Optimize for response quality. 50 emails with sharp personalization > 500 generic.
- Don't add complex CRM integrations or pipeline tools yet. Folk + Cal.com is enough until ~10 customers.
- Don't draft op-eds in the voice of a SaaS founder. The [INDUSTRY] trade press has a specific register — analytical, regulation-aware, references the domain's specifics by name. Match it.
Useful references when editing here
- Your master plan — the 90-day phase plan and verification criteria.
<!-- TODO: link your master plan --> pipeline/build.md— process for moving a row from raw → enriched → scored → outreach-ready.metrics/weekly.md— what numbers to actually look at on Friday.
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.

