Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Skill Listing Writer: Marketplace Copy Backed by Your Files

    1

    Draft a skill listing from what the package actually contains: name, description, tags, and a claims register that shows which inventory value each claim matched, for you to confirm.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Skill Listing Writer: Marketplace Copy Backed by Your Files

    Skill Listing Writer: Marketplace Copy Backed by Your Files

    Example session with this skill installed

    Draft an Agensi listing for my skill in devils-advocate-review-panel.zip. Inventory the package first, check the draft's claims against it, and check that the display name starts with "Review Panel".

    • Read your context and instructions
    • Compiled the listing writer

    What happened

    The skill inventoried the ZIP, the agent drafted the fields from that inventory, and the scripts checked the draft. Real output from the bundled scripts (v1.2); the inventory JSON is summarised:

    $ python scripts/inventory.py devils-advocate-review-panel.zip > inv.json

# inv.json keys: root, source, frontmatter, headings, scripts, file_types, imports

$ python scripts/claim_check.py draft.md inv.json --profile assets/profiles/generic.json
Claims found: 3

  1. [ ] [term_match] panel_plan.py generates role cards for each review round.  — matched: scripts (panel_plan.py): "Generate role cards for a multi-round review panel."
  2. [ ] [term_match] findings_ledger.py merges findings and evaluates a gate per round.  — matched: scripts (findings_ledger.py): "Merge and evaluate findings from a multi-round review panel."
  3. [X] [term_match] **Summary:** Multi-round QA gate that finds all errors in any deliverable.  — matched: frontmatter (SKILL.md): "description: "Use when you need a structured multi-round review of any deliverable (text, plan, code, listing, or report"
       PROMISE: promise word: 'any'
       PROMISE: promise word: 'all'

Evidence: 3 term matches to confirm, 0 not verifiable, 0 no evidence (of 3 claims)
Issues found: 1
  - Claim 3: ["promise word: 'any'", "promise word: 'all'"]
(exit 1)

$ python scripts/headword.py "Devil's Advocate Review Panel: Multi-Round QA Gate" "Review Panel"
FAIL: "Devil's Advocate Review Panel: Multi-Round QA Gate" does not start with 'Review Panel'
Suggestion: rename to 'Review Panel: Devil's Advocate Review Panel: Multi-Round QA Gate'
(exit 1)

    The claims register, after you confirm

    ClaimMatched valueYour decision
    panel_plan.py generates role cards for each review round.scripts (panel_plan.py): "Generate role cards for a multi-round review panel."confirmed
    findings_ledger.py merges findings and evaluates a gate per round.scripts (findings_ledger.py): "Merge and evaluate findings from a multi-round review panel."confirmed
    Summary: Multi-round QA gate that finds all errors in any deliverable.frontmatter (SKILL.md): "…a structured multi-round review of any deliverable…"remove "all errors" and "any"; the matched text does not back that promise

    Next steps

    1. Rewrite the summary without "all" and "any".
    2. Rename the listing so it starts with the headword, for example "Review Panel: Devil's Advocate Multi-Round QA Gate".
    3. Re-run the claim check, confirm or qualify each term_match, then paste the fields into the marketplace form.
    frontmatter (SKILL.md

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Use when you have finished building an agent skill and need marketplace copy grounded in what the package actually contains.

    How it works:

    1. The skill's scripts inventory the folder or ZIP: frontmatter, headings, scripts, argparse flags and imports.
    2. Your agent writes the listing fields from that inventory only.
    3. A claim check flags promise words and builds the claims register: each claim with its status (term_match, not_verifiable or no_evidence) and, for a term match, the inventory value it hit (key and file), plus a tally line.
    4. You confirm, qualify or remove each term_match. A match shows where to look; it does not prove the claim.
    5. A headword check confirms that the display name starts with your headword.

    Do not use when you are writing a skill for end users to interact with; this is for skill authors preparing upload copy. Do not use when your skill files are not yet readable on disk or inside a ZIP.

    Requires: Python 3.9 or later (standard library only).

    Limits: Static analysis only; does not run your code. Does not report installs, sales, approval status or rankings. Matching is by shared terms, so an unrelated claim can still produce a term_match; read the quoted value before you keep the claim. The promise-word check is a word list, so a reviewer should still read the draft.

    Independent: Not affiliated with or endorsed by Anthropic, OpenAI, Anysphere (Cursor) or Google. Product names are used only to describe compatibility.

    Static analysis only; does not run your code. Does not report installs, sales, approval status or rankings. Matching is by shared terms, so an unrelated claim can still produce a term_match; read the quoted value before you keep the claim. The promise-word check is a word list, so a reviewer should still read the draft.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 8 days ago

    Requires

    Python 3.9+ (standard library only)

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Listed8 days ago

    What's inside

    Frequently Asked Questions

    No. All three scripts are static and use only the Python standard library.

    No. Field limits come from a platform profile JSON you supply; a generic profile is included.

    It is flagged as not verifiable, and the skill tells you to remove it or mark it as a heuristic.