agentleFS
calven-aicalven-ai/marketing-as-code/.agents/skills/campaign-discovery/SKILL.md

campaign-discovery

Skill10 starsChanged 40 days agoMITView on GitHub

Discovery for a campaign idea in one sentence: competitive angle, keywords and ranks, AI answer coverage, existing content, one report. Use when "explore a campaign", "run discovery on".

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 1 is distinctive, and 5 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.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Competitive angle: Read strategy/competitive/ (the battlecards) for who competes on this idea and how they position.

    line 34
  2. Very specific:

    Keywords: Ask seo-analyst for volumes, difficulty, and our current rank for the core terms plus up to ten ideas.

    line 39
  3. Very specific:

    AI answer-engine coverage: Ask brand-monitor to run the prompts in data/seo/prompts.csv that touch this idea (and to propose up to three …

    line 43
  4. Very specific:

    Content inventory: Run content-inventory and take its report: what already exists that this campaign could reuse, refresh, or must not …

    line 46
  5. Distinctive:

    Every number traces to a snapshot.

    line 85
  6. Very specific:

    Do not scaffold the project or any content from here.

    line 87

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Campaign discovery

One sentence in ("a campaign positioning us as the AI-native alternative for marketing ops teams"), one discovery report out. You compose the specialist skills; you do not decide whether the campaign happens.

Needs: strategy/ filled (positioning, messaging, competitive/). The keyword part needs a wired seo-data integration and the answer-engine part a wired ai-visibility one; which vendor fills each is the Wired table in integrations/README.md, and seo-analyst and brand-monitor resolve it. Both are optional: without them, run the parts that work offline and mark the rest as gaps, naming the export a person could drop into data/seo/snapshots/YYYY-MM-DD-<vendor>-<what>.csv.

Procedure

  1. Frame the idea. Restate the campaign in one line, name the persona from strategy/personas.md and the messaging pillar from strategy/messaging.md it leans on. If neither fits, that is the first open question.
  2. Competitive angle. Read strategy/competitive/ (the battlecards) for who competes on this idea and how they position. Ask seo-analyst for the SERP competitors on the campaign's two or three core terms, so the report lists both the known competitors and whoever actually owns the search results. Note the angle they leave open.
  3. Keywords. Ask seo-analyst for volumes, difficulty, and our current rank for the core terms plus up to ten ideas; the pull lands in data/seo/snapshots/. Flag which terms are in keywords.csv already and which would be proposed additions.
  4. AI answer-engine coverage. Ask brand-monitor to run the prompts in data/seo/prompts.csv that touch this idea (and to propose up to three new prompts for it). Report who is cited today and whether we are.
  5. Content inventory. Run content-inventory and take its report: what already exists that this campaign could reuse, refresh, or must not duplicate, grouped by channel with status and owner.
  6. Write the report to reports/adhoc/YYYY-MM-DD-<slug>-discovery.md from reports/_templates/report.md, with these sections after the answer: competitive angle, keywords (table), AI answer-engine coverage (table), content inventory (by channel), what we would need to produce (a list: piece, channel, why, reuse or new), open questions for the team, and Data used with every snapshot path.

Worked example

"Run discovery on a campaign about meeting transcripts turning into tasks automatically."

  • Framing: persona "marketing operations lead"; pillar "meetings stop evaporating".
  • Competitive angle: two battlecards mention note-taking tools; SERP competitors for "meeting notes to tasks" are three note-taking apps, none positions on the marketing team. Open angle: the whole team's memory, not one person's notes.
  • Keywords: "meeting transcript to action items" 1,900/mo, difficulty 34, we do not rank; "ai meeting notes" 40,500/mo, difficulty 71, not realistic. Snapshot data/seo/snapshots/2026-09-03-dataforseo-keyword-overview.csv.
  • AI coverage: for "Which platforms turn meeting transcripts into decisions and tasks?" two note-taking apps are cited, we are not.
  • Content inventory: one published blog post on decision logs (reuse), one LinkedIn draft in review (refresh), no email.
  • What we would need to produce: a guide on the transcript pipeline (blog, new), an email to existing users (email, new), a refreshed LinkedIn post (reuse).
  • Open questions: is the operations lead the buyer or the user; do we have a customer quote we can use.

Rules

  • Propose, never decide: the report ends with questions and a list, not a go / no-go.
  • Every number traces to a snapshot; a skill you could not run is a named gap.
  • Do not scaffold the project or any content from here; new-project and new-content are separate, human-triggered steps.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.