grow-seo
Skill1 starsChanged 3 months agoMITView on GitHub
Build organic search as a channel — technical audit, keyword-to-funnel mapping, competitive read, prioritized fixes, and content briefs handed to /create-content. Run once to establish the plan, then quarterly to refresh; rank tracking rolls into /measure.
In this repository22 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- build-funnelSkill
- close-dealSkill
- connect-toolsSkill
- create-contentSkill
- define-positioningSkill
- design-offerSkill
- experimentSkill
- grow-referralsSkill
- health-checkSkill
- launch-campaignSkill
- launch-productSkill
- manage-pipelineSkill
- measureSkill
- plan-strategySkill
- research-marketSkill
- review-monthlySkill
- run-adsSkill
- run-outboundSkill
- run-partnershipsSkill
- setup-teamSkill
- tune-revopsSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 16
Build the question list: Start from knowledge/research/customer-language.md, expand with the Ahrefs MCP (or Bright Data search), and record …
- Very specific: line 32
No site to optimize (marketplace-only or no web presence) → run the whole method against the marketplace listings.
- Distinctive: line 33
Every candidate term reads as closed → that's a real finding, not a failure.
≈ “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
/grow-seo — The Channel That Compounds
Inputs: knowledge/positioning.md, knowledge/audience-profiles.md, knowledge/research/customer-language.md, knowledge/competitors.md, knowledge/strategy.md, the site or listings themselves.
Agents: seo-specialist leads; tools-operator runs Ahrefs / Bright Data seo-audit; content-marketer receives the briefs; funnel-builder fixes technical and on-page issues.
Method: knowledge/methods/seo.md.
Steps
- Set expectations first. State the timeline honestly: 4–8 months to meaningful movement. If the business needs sales sooner, say so and position SEO as the compounding investment running alongside the faster channel — never as the answer to an urgent revenue gap.
- Audit what exists. Technical health (indexation, speed, mobile, canonicals, broken links, sitemap) plus current rankings and any pages already earning impressions. Marketplace businesses: audit the listings instead — titles, tags, images, review volume.
- Build the question list. Start from
knowledge/research/customer-language.md, expand with the Ahrefs MCP (or Bright Datasearch), and record volume and difficulty for each — weighting difficulty more heavily. - Read the results pages. For every serious candidate, examine what actually ranks: page type, thoroughness, freshness, site authority. Mark each term open or closed, and say why. Closed doors get abandoned here, not discovered six months later.
- Map topics to funnel stages. Group related questions into clusters; assign each to AWARE / TRUSTING / BUYING per
playbook.md§1. Prioritize commercial intent — it converts far better and is less contested. - Prioritize the work. Rank every opportunity by effort vs. likely return. In practice the order is nearly always: (a) refresh pages already on page two, (b) fix anything technically broken, (c) improve titles and descriptions where impressions exist but clicks don't, (d) internal linking, (e) new content — last, because it's slowest.
- Write the briefs. For each prioritized topic, fill
templates/content-brief.md— target question, intent, what ranks and what it misses, the angle, must-cover points, structure, internal links, conversion path. Hand to/create-content; the calendar picks them up ahead of invented topics. - Instrument and set the review date. Confirm Search Console access and organic tracking so
/measurecan report impressions and average position monthly. Judge content at 3 months, again at 6 — record those dates now.
Before you start
- Required inputs:
knowledge/positioning.mdandknowledge/audience-profiles.mdatseeded+;knowledge/research/customer-language.mdpresent (without it, keyword research becomes guessing at phrasing). - Scan
knowledge/decision-log.mdfor tagschannel,content— if a prior decision deprioritized SEO, this run either honors it or supersedes it explicitly. - Gate check: none for research and briefing. Site changes are public → publishing gate via
/build-funnelor/create-content.
If things go wrong
- Ahrefs unauthorized → use Bright Data
seo-audit/searchplus Search Console; note that difficulty scores are estimates rather than measured, and say which numbers came from where. - No site to optimize (marketplace-only or no web presence) → run the whole method against the marketplace listings; the discipline transfers unchanged. If there's no surface at all, recommend
/build-funnelfirst. - Every candidate term reads as closed → that's a real finding, not a failure. Report it, and recommend the channel where this audience is actually reachable.
- Site is technically broken beyond quick fixes (platform limitations, no access) → prioritize what's fixable, park the rest in
knowledge/pending.mdwith exact instructions.
Outputs
An SEO audit with prioritized fixes, a keyword-to-funnel topic map, and a set of content briefs queued into the calendar. Close with the report per templates/skill-report.md.
Writeback
knowledge/research/seo-<YYYY-MM>.md (audit + topic map); briefs into knowledge/content-calendar.md; technical fixes to knowledge/decision-log.md (tag: content); tracking setup to knowledge/tool-registry.md; a learning line to knowledge/methods/seo.md once results arrive.
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.

