agentleFS
gitTerebigitTerebi/seo-content-machine-plugin/skills/scm/SKILL.md

scm

Skill0 starsChanged 26 days agoMITView on GitHub

Use SEO Content Machine (SCM) through its MCP tools to write AI pages from page templates, scrape articles from search results, crawl a site, list search results, pull Google Maps business listings, save YouTube subtitles, download archive.org copies of a site, extract values from pages into CSV, or plan and write pages in an SEO Workspace. Use when the user mentions SCM or SEO Content Machine, or asks for one of those jobs.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. first-fluke/fullstack-starter★ 226same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repositoryOne other file it gives its agents.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 1 is distinctive, and 4 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.

  • 4

    Very specific

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

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

    Call task_list first in a new session.

    line 14
  2. Very specific:

    Every *_create tool and page_pipeline_upsert only make a draft task.

    line 18
  3. Distinctive:

    Create it. The reply has the task id.

    line 20
  4. Very specific:

    Start it with task_run and that id.

    line 21
  5. Very specific:

    Do not start many large tasks at once.

    line 69

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

SEO Content Machine

SEO Content Machine (SCM) is a desktop app. Its MCP tools create and run tasks inside the app, which does the browsing, scraping, writing and file saving on the user's machine. Results land in the task's output folder on that machine.

Before the first call

The app must be open. If no SCM tools are available, or a tool answers "Cannot reach SCM", ask the user to start SEO Content Machine and log in. The tools appear on their own within a few seconds of SCM starting. Do not retry in a loop.

Call task_list first in a new session. It shows the tasks that already exist and how busy the machine is.

How a task runs

Every *_create tool and page_pipeline_upsert only make a draft task. The pattern is always:

  1. Create it. The reply has the task id.
  2. Start it with task_run and that id.
  3. Check progress with task_list filtered by the task's name. It returns the status and run counts without the task's settings. Call task_get with include_logs: true only to read the log, once the run ends or when it looks stuck or failed, because task_get also returns every setting.
  4. Call task_results. Its rows point to what the run wrote: a file, files, a CSV or a sheet, and run.output_folder is the output folder. Read those files with your own file tools. A Maps CSV with review text is wide, so read only the columns you need.

Runs take from seconds to many minutes. Poll task_list every 20 to 60 seconds rather than in a tight loop, and tell the user what is happening while you wait.

task_run resumes an unfinished run. task_run with fromStart: true starts over from the first input line. task_cancel stops a run.

Which tool for which job

JobTool
AI pages from page templates, or articles scraped from search results, for a list of keywordspage_pipeline_upsert
Write pages already planned in an SEO Workspaceworkspace_create_pages
Specific values from pages (prices, schema, titles) into CSVdynamic_pages_create
Whole pages saved as cleaned article files, no browserstatic_pages_create
Every internal link on a sitesite_crawler_create
Search results (rank, URL, title) for a list of queries, as one CSVurl_finder_create
Subtitles of YouTube videos, by URL or keyword searchyoutube_subtitles_create
A dead or old site rebuilt from archive.orgweb_archives_create
Businesses in a place, with ratings and reviewslocal_listings_create
A quick web searchweb_search
Read one page after its JavaScript runsbrowser_open
Click, type or log in on a sitecdp_open
Print HTML to a PDFrender_pdf

For "best", "top rated" or "near me" questions about businesses in a place, use local_listings_create with a line such as "plumbers in denver", not a web search.

Output folders accept %project_folder%, which is the task's own project folder.

SEO Workspace pages

An SEO Workspace is a task of type "seo workspace". Find its id with task_list and type: "seo workspace", or make one with seo_workspace_create. That id is the task_id the workspace tools take.

  • workspace_query reads one collection, such as pages, sitePlan, brief or documents, with a Mongo filter. Pass fields to keep the reply short.
  • workspace_write inserts or updates rows straight away, with no confirmation step. Show the user the list of pages and get a yes before you add them. To update exact pages, filter with {"pageId": {"$in": [...]}}. Page and Site plan changes are journaled, and the user can undo them in the app.
  • workspace_create_pages hands pages to a Page Pipeline, which researches, writes and reviews one page each and saves it as a draft document. Pass run: false to set the model and effort with page_pipeline_upsert before task_run. Use effort "medium" or higher, because on low the writer stops researching.
  • workspace_documents_edit replaces exact text in one document's body.

For local service pages across towns, follow the local-seo skill.

The user reviews and publishes pages in the app. Do not describe a page as published because its status changed.

Care

  • task_delete removes the task folder, including its results and log. Confirm with the user first.
  • AI images in a Page Pipeline cost SCM AI credits (5k per image on the scm service). Mention the cost before turning them on.
  • AI pages use the user's own AI service, roughly 75k to 300k tokens a page. Mention it before a large batch.
  • Do not start many large tasks at once. task_list shows the machine's CPU, RAM and task load.

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.