Dobeu-tech-eco

gtm

"Take a product through go-to-market in one pass: installs marketing-skills if missing, runs a short GTM deep-dive and the skills that apply, and ends with a launch-readiness scorecard. Use for go to market, launch readiness, gtm, or SEO review."

Dobeu-tech-eco 0 Updated 1w ago
GitHub

Install

npx skillscat add dobeu-tech-eco/dts2-marketplace/gtm

Install via the SkillsCat registry.

SKILL.md

Directory edition: do not run bundled scripts or hooks, a package launcher, or a download piped into a shell. Do not read tokens or credential files from the machine. Carry out the step with Read, Edit, Grep, Glob, and git or gh, and write the files the step names.

GTM — one seamless pass

This command never hard-stops. If something is missing it installs it, falls back, or continues with what is available.

1. Ensure marketing-skills (never stop)

  • Already available? Check with /plugin or ListPlugins, or by whether the Skill tool can invoke marketing-skills:<name>. If so, use it.
  • Not available: install headlessly with the publisher's installer for https://github.com/coreyhaines31/marketingskills (the same headless mechanism the harness uses for its companion skills), then use the installed skills.
  • Only if that genuinely cannot run (offline, or a non-synced environment): print ONE line — enable marketing-skills@synced under synced plugins, or run /plugin install marketing-skills@claude-cowork — and continue with whatever is available.
  • Prefer the marketing-skills skills and do not hand-reimplement them when they are available.

Run python3 "${CLAUDE_PLUGIN_ROOT}/scripts/ensure-ignored.py" <repo>.
Dispatch dt:scout for the marketing-surface list; without .harness/end-state.md, say so and continue with scout's output only.
A scout report whose first line is PARTIAL: hit turn cap at <step> is not the surface list: re-dispatch the scout with a narrower scope or a fresh agent, inside the fix-loop cap.

2. GTM deep-dive

Before the pipeline, run a short deep-dive.

  • Read .harness/end-state.md and .harness/latest.md first. Do not re-ask anything they settle: the rebrand direction is already approved, so treat it as fixed.
  • Ask one to three clarifying questions, only on what is still open: positioning angle, primary audience, launch goal and date, priority channels.
  • Offer to re-run the deep-interview project parse (the end-to-end sectioned parse) to feed GTM. If the user declines or is away, proceed from the existing end state.
  • Write the answers into .agents/product-marketing-context.md as inputs.

3. Select the skills

Run python3 "${CLAUDE_PLUGIN_ROOT}/scripts/marketing-select.py" --facts '<json>' with facts from scout and the deep-dive (has_web, has_landing, has_marketing_surface, analytics). It returns the ordered subset of marketing skills this repo needs. The full pipeline, in canonical order, is below; selection chooses which of these run, never their order.

4. Run as one pass

Execute the selected skills back to back, feeding each the end state, the marketing-surface list, and .agents/product-marketing-context.md. Narrate it as a single GTM pass: do not announce each skill or present them as separate skills to the user. Each still writes its report behind the scenes.

Canonical order (the full pipeline):

  1. /marketing-skills:product-marketing-context — auto-draft from the repo.
  2. /marketing-skills:competitor-profiling
  3. /marketing-skills:seo-audit
  4. /marketing-skills:page-cro
  5. /marketing-skills:copywriting
  6. /marketing-skills:copy-editing
  7. /marketing-skills:launch-strategy
  8. /marketing-skills:analytics-tracking — runs only when selected (analytics fact true).

5. Outputs

Write to .harness/gtm/<stamp>/ (<stamp> = YYYYMMDD-HHMM):

  • one report per step that ran, named <step>.md (for example seo-audit.md);
  • readiness.md: a table with columns area | ready (yes/no) | blocker | owner, one row per step plus any gap the steps surfaced.

Finish by showing readiness.md and naming the top blocker, as one unified result.