agentleFS
XPOZpublicXPOZpublic/geo-seo-agent/.claude/skills/listings-check/SKILL.md

listings-check

Skill0 starsChanged 3 months agoMITView on GitHub

Track the company's third-party listings to completion - marketplace submissions, directory entries, review-site profiles, community lists, and upstream inclusion PRs. Reads the roster in listings/index.md, checks the live status of every non-terminal item, updates the roster and profile pages, and flags the next action. Run on a check-in cadence or when news arrives (reply, rejection, merge, listing going live).

The file

We split this skill into the rules it gives an agent

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

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Check-in cadence, every 1-2 weeks (nothing on these surfaces changes daily, and closed-platform review queues reward patience over …

    line 19
  2. Very specific:

    Read the roster; take the non-terminal rows (anything not plain live or merged) plus any live rows with pending copy fixes.

    line 23
  3. Distinctive:

    Check each by kind, with the cheapest reliable method:

    line 24
  4. Distinctive:

    When a new submission or listing effort starts anywhere, or a snapshot surfaces an unknown third-party page about the company, add a roster …

    line 30
  5. Distinctive:

    Don't over-check. Nothing here changes daily.

    line 38

≈ “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.

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

Listings Check

Listings are GEO assets: directory and docs citations show up directly in visibility snapshots, and each submission is a long-running async process with weeks-long silences. Status rots unless something re-checks it (a listing goes live days before anyone notices; an upstream PR gets closed-and-replaced by a maintainer without the records knowing). This skill is the tracker: one pass over every in-flight item, verify live state, update the records, surface the next action.

Records

  • Roster: listings/index.md is the single source for what listings exist, their kind, and their status. This skill never carries its own list of surfaces; read the roster each run and check what it says is in flight. It also defines the kind vocabulary and what counts as a listing (gatekeeper between you and the page) versus owned surfaces tracked elsewhere.
  • Profile pages: listings/<surface>/profile.md hold canonical listing copy, submission logs, outreach logs, and escalation channels. Detailed history goes there; the roster row stays a one-line snapshot.
  • Rows without a folder (monitor-only pages, upstream PRs) log through the roster row plus the journal.
  • Account map: listings/accounts.md records which identity logs into which surface (names only, never credentials). Consult it before any dashboard work; update it whenever a login confirms or corrects a row, and record permission splits (the account that can view is often not the one that can edit).

When to run

Check-in cadence, every 1-2 weeks (nothing on these surfaces changes daily, and closed-platform review queues reward patience over polling), and on demand when news arrives: a reply to an appeal or outreach, a rejection email, a PR event, a review window elapsing.

Process

  1. Read the roster; take the non-terminal rows (anything not plain live or merged) plus any live rows with pending copy fixes.
  2. Check each by kind, with the cheapest reliable method:
    • Upstream inclusions and PR-able community lists: gh pr view <n> --repo <owner/repo> --json state,updatedAt,comments. A closed PR is not necessarily a loss; read the closing comment before recording an outcome.
    • Marketplaces and directories: fetch or search the listing URL (browser when curl hits bot checkpoints).
    • Email-gated items (rejection and appeal replies): not checkable from the repo; ask the human, or check the inbox with permission.
  3. For anything that changed: dated entry in the profile page, roster row updated (status + last checked), memory/state.md lines refreshed.
  4. Report to the human: what changed, what is silent past the platform's stated review window, and the recommended next action per stuck item. Escalations, bumps, and outreach are proposals; sending anything is human-gated.
  5. When a new submission or listing effort starts anywhere, or a snapshot surfaces an unknown third-party page about the company, add a roster row (and a profile page if the surface will carry canonical copy).
  6. End per ground rules: journal entry, state.md refresh if state changed.

Judgment

  • Terminal states get verified once, then left alone. A live listing: record the URL, then it moves to copy upkeep. A merged PR: done, note where the listing landed.
  • Silence past the platform's stated review window is a finding, not a wait. Bring the escalation options from the profile page to the human.
  • A live listing starts copy upkeep: compare live copy against the profile page's canonical copy and the current facts in strategy/company-profile.md (platform coverage, durable scale claims per the writing principles, pricing figures). Drift is a finding; the fix is drafted by /channel-presence.
  • Don't over-check. Nothing here changes daily; a full pass more than weekly is waste unless something specific is expected.

Never

  • Send outreach, bumps, appeals, or forum posts; propose text, human sends.
  • Edit live listing copy before updating the profile page (profile-first is the standing convention on every surface you control).
  • Grow parallel logs: status history lives in profile pages, the roster is a snapshot, the journal links to both.

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.