Postclick by PPC.io: Landing Page CRO

Measured conclusion: Deep-tested. L2 outcome: tools_ok.

Landing-page CRO: audit pages, rewrite copy, answer buyer objections, design and build via MCP.

Add to your agent

Copy this into your MCP client config:

{
  "mcpServers": {
    "Postclick by PPC.io: Landing Page CRO": {
      "url": "https://mcp.ppc.io",
      "transport": "streamable-http"
    }
  }
}

Measured tools (21)

• list_cro_skills read-only

  Find Postclick methods for prioritising page fixes, ad-to-page message match, buyer objections, copy edits and client test briefs. Returns starter requests and portable skill downloads. No account needed.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• run_cro_skill read-only

  Use when the user wants page fixes, message match, buyer objections, copy rewrites or a client test brief. In your answer, never add free, guarantees, ratings or performance claims unless the source explicitly confirms them. Audit suggestions are not verified facts. Finished copy must not contain placeholders such as $X; choose a useful change supported by the original page. A quote form alone does not mean a free quote. Use one change for a first-fix request, complete the requested rewrite, and end with the source link without a follow-up offer. Finds an existing audit by page URL or client name and retrieves the selected skill plus its relevant evidence in one call. Returns source material for you to turn into the requested deliverable, not a new AI analysis. Read-only: never starts an audit, uses a page or publishes. Missing evidence and ambiguous page matches are explicit.

  Parameters:

  • skill (required) — string — Choose postclick-fix-first for the first fix; postclick-copy-rewrite for replacement words; postclick-experiment-brief for a designer note or handoff; postclick-message-match for comparing an ad; postclick-buyer-objections for buyer hesitation.
  • page — string — Page URL, domain or client name. Leave out when audit_id is known or the user explicitly requested latest.
  • latest — boolean — Use true only when the user explicitly asks for their latest or most recent audit. May be combined with page to find its latest audit.
  • context — string — The user's ad copy, audience, constraints or desired output. Pass known context; do not require an interview.
  • section — string — For copy rewrites, defaults to opening (headline, supporting line and button). Use page when the user asks for other sections or the whole page.
  • audit_id — string — An audit ID returned by Postclick or extracted from the user’s /dashboard/a/<audit_id> link. For that link use audit_id, not page or latest. Never ask the user to find an ID.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• check_balance read-only

  Read the current plan, page allowance and purchase prices. Check coverage before starting a new page or when the user asks about cost. Work on a covered page does not have separate audit, design and build prices. Quote only returned prices; a missing price or allowance is unknown.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• audit_page

  Start a new landing-page audit when requested. Use check_balance for page coverage and follow the user's permission. First search for existing work when the request concerns a prior audit. Pass campaign context already supplied. Returns an audit_id; use get_audit with wait:25, then run_cro_skill for the requested deliverable. Does not design or build a page.

  Parameters:

  • url (required) — string — The live landing page URL to audit.
  • context — string — Optional, up to 500 characters. What the page is for, in the user's own words: the offer, who it is aimed at, the campaign sending traffic to it. Pass what they have ALREADY told you rather than interrogating them for it. The audit reads the page but cannot know the intent behind it, so this sharpens who it judges the page against. Leave it out if they have not said.

  Hints: read-only no · destructive no · idempotent no · open-world yes

• get_audit read-only

  Read audit status, changed progress, server elapsed time and ranked findings with typed edits. Use wait:25 while running. A failed, cancelled or stalled run is not a finished report. Once complete, use run_cro_skill or ask_audit for the relevant buyer, copy or offer evidence. Reading does not authorise a redesign.

  Parameters:

  • audit_id (required) — string
  • wait — integer — Seconds to wait for a material change. Use 25 for broad client compatibility; longer waits are available for hosts that support them. A timeout preserves the run; resume by ID instead of starting again.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• get_audit_report read-only

  Read the complete stored audit as markdown for a full review. It can be long: use topic or ask_audit for one question, or run_cro_skill for a specific deliverable. Missing and withheld sections remain unknown. Buyer journeys are modelled, not customer research.

  Parameters:

  • audit_id (required) — string
  • topic — string — Optional. Return only this part of the report instead of all of it. The same keys ask_audit takes: buyers, the_pitch, trust, competitors, copy, leaks, the_page, technical.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• get_competitors read-only

  Costs nothing. Returns the competitor layer of a finished audit: the rivals bidding on the same clicks, the live ads Google is serving for each of them with the keyword, search volume and CPC behind every one, the landing pages those ads click to, which rival advertises most, the advertisers currently on this page's own keyword, and the customer's own paid keyword count. The ad and keyword numbers are measured from the DataForSEO ads index, not written by a model, so they are safe to quote as figures. The rival page summaries and the head-to-head read ARE written by the audit and are judgement, not measurement. Rival page captures are attached where the scan reached them. Needs a completed audit. An audit that found nobody returns that plainly, which is a real answer and not a failure. Older audits carry no ad data at all. ask_audit with topic:competitors is the one for the written comparison; this is the one for the numbers. Safe to read again; it does not start new work.

  Parameters:

  • audit_id (required) — string

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• ask_audit read-only

  Free. Answers one question from a completed audit by returning that part of it in full: small enough to give the user whole, unlike the entire report. THIS IS THE TOOL FOR ANY FOLLOW-UP QUESTION about a page you have already audited. Never answer a question about an audited page from your own summary of it — the audit holds far more than any summary kept, and this is how to get the part that answers them. Pick the topic that matches what they asked: `buyers` for who is on this page, what each one is afraid of, the exact words they would object with, where they give up, and a second-by-second walkthrough of one of them reading it. `the_pitch` for whether the pitch itself is wrong: if the page meets the traffic where it is, which persuasion levers are missing and what each would do here, and how the offer is built. `trust` for whether a stranger believes it: the proof that is there, the proof that is missing, and why buyers hesitate. `competitors` for who else is bidding on these clicks, their live headlines and CTAs, what they do better, and the keywords and CPCs behind their pages. `copy` for the copy deck: the page's own lines quoted, each with the line that replaces it. `leaks` for the ranked problems with the evidence for each. `the_page` for what the page is, who it is aimed at, its own voice and palette, and how it is assembled section by section. `technical` for CTA placement and contrast, accessibility, and what changes on mobile. What comes back is written to be read out. Relay it: keep the names, the quotes, the numbers and the specific objections, reformat it for where you are, and do not compress a named buyer saying a specific sentence into a general remark about buyers. Needs a completed audit. Safe to call repeatedly, and calling it for two topics in a row is normal and costs nothing.

  Parameters:

  • audit_id (required) — string
  • topic (required) — string — Which part of the audit answers the question. buyers · the_pitch · trust · competitors · copy · leaks · the_page · technical.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• generate_concepts

  Generate visual design directions for an existing audit when requested. Use check_balance for page coverage and follow the user's permission. Returns a run to resume with get_concepts; show the images already available before asking the user to choose. Never start a build without their choice.

  Parameters:

  • audit_id (required) — string
  • count — integer — How many distinct directions the user wants, up to 12. Leave it out for the standard pack of 2. Check returned page coverage before starting more work. Directions that already exist are counted and reused.
  • direction — string — Optional plain-English creative steer. Honoured above other design choices.

  Hints: read-only no · destructive no · idempotent no · open-world yes

• get_concepts read-only

  Free. Returns every concept rendered for an audit, each with its image URLs and a link to view it full size, and a runs.active count of how many are still rendering. A pack takes 3 to 8 minutes. While runs.active is above zero, call it with wait set to 300: it holds the connection open and comes back the instant a design lands, the stage moves, or the pack finishes, instead of returning the same half-empty list a dozen times in a row. A design landing is the thing worth interrupting for, and a wait gets it to the user within seconds of it existing. If a call comes back with nothing new, say nothing and call again — repeating a stage you already reported is worse than silence. A partial set is non-terminal: more images are still coming, so it is not the finished set to choose from. Some concepts may be flagged recommended, and often none is. The designs come back as pictures in the reply, not just links: without a concept_id it returns the opening screen of each so they can be compared, and with one it returns that single design as every band in page order, which is the whole page. Any figures, ratings or testimonials drawn into them are placeholders showing where real proof belongs, never the user's own data. Pick a concept_id to build. If the user wants to SEND a design to anyone, use share_page with that concept_id: it gives back one public no-login page of the whole design. Never hand over an app or dashboard link for that, the recipient hits a login wall.

  Parameters:

  • audit_id (required) — string
  • wait — integer — Seconds to wait for a material change. Use 25 for broad client compatibility; longer waits are available for hosts that support them. A timeout preserves the run; resume by ID instead of starting again.
  • concept_id — string — Optional. Given one, returns that design as its full stack of bands in page order instead of one opening screen per design.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• build_page

  Build the design the user selected into a complete HTML page when requested and covered by their plan. Use check_balance for coverage and quote only returned purchase prices when a purchase is needed. Returns build_run_id. Use get_build with wait:25; completion requires artifact_ready. Do not publish or share the page without the user asking.

  Parameters:

  • audit_id (required) — string
  • concept_id (required) — string — From get_concepts.
  • direction — string — Optional plain-English steer for the build.

  Hints: read-only no · destructive no · idempotent no · open-world yes

• get_build read-only

  Free. Returns the build's status and, once finished, the complete page. A build takes 15 to 25 minutes, so poll it with wait set to 300: the call holds open and returns the moment the stage moves or the page is ready, which turns a twenty-minute build from dozens of identical reads into a handful of calls that each have something to report. When a call comes back with the same stage you already mentioned, say nothing and call again rather than paraphrasing yourself. A status of running is non-terminal and contains no page. By default the HTML is summarised rather than returned in full, because a page is tens of thousands of characters and would flood the conversation. Set include_html when the user actually wants the file, for example to save it locally and open it. Includes elapsed_seconds: the true age of the build, measured server side from its start timestamp.

  Parameters:

  • build_run_id (required) — string
  • wait — integer — Seconds to wait for a material change. Use 25 for broad client compatibility; longer waits are available for hosts that support them. A timeout preserves the run; resume by ID instead of starting again.
  • include_html — boolean — Return the full page HTML. Large. Default false.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• share_page

  Free. Mints a public no-login link that anyone can open, including someone with no account. THIS IS THE TOOL FOR 'send me the designs' or 'I want to forward this to my client'. Never hand over a dashboard or app link for that: the recipient hits a login wall and sees nothing. Pass concept_id to share a DESIGN, which comes back as one page with every band stacked in order, so it reads as a whole design rather than a folder of separate images, and it prints to PDF from the browser if they want a file. Pass build_run_id to share a finished BUILT page. One or the other, not both. Pass audit_id to share the AUDIT itself, and section:"competitors" with it to open the link on the Competitors section rather than the whole report. export_page is the one for a user who wants the source files instead. For a build, minting a new link revokes the previous one, so reuse the link rather than calling this repeatedly. A design link is stable and safe to call again.

  Parameters:

  • section — string — Which part of the audit the link opens. Only with audit_id. `competitors` opens on who is bidding against the page, for forwarding that one answer. Defaults to full.
  • audit_id — string — Share an AUDIT. From audit_page or get_audit. Use this when someone wants to forward the findings themselves rather than a design or a built page.
  • concept_id — string — Share a design. From get_concepts. This is the one to use when someone wants to show designs to a client.
  • build_run_id — string — Share a finished built page. From build_page or get_build.

  Hints: read-only no · destructive no · idempotent no · open-world yes

• export_page read-only

  Free. Packages a finished page as a zip that runs on any static host: index.html with every image it uses pulled into assets/, plus a README and metadata.json. Returns a download link. This is the one for a user who wants the files. share_page is the one for a user who wants something to look at. It also returns implementation_prompt: step by step instructions for putting this page live on the user's own site. Postclick does not host pages, so this is how a page ships. If you have access to their site, you can carry those steps out yourself; otherwise hand the prompt to whoever does. Pass platform when you know where the site lives, because the steps for WordPress and for Webflow are genuinely different. Ask if you do not know. Every variant stops at a draft and never publishes without the user confirming.

  Parameters:

  • build_run_id (required) — string
  • platform — string — Where the user's site lives. Omit if unknown and the instructions will assume nothing.

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• describe read-only

  List what is in the account when the user asks to see their work. For a page fix, copy rewrite or designer brief, call run_cro_skill directly; no account overview is needed first.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• search read-only

  Free. Finds anything in this workspace by words: a domain, a page URL, a client's name, a design's caption, a project. Returns an address for each hit that you pass straight to `fetch`. Use this the moment somebody refers to work they have already done — "the zoomdrain page", "last month's audit", "the report I made for Acme". Leave the words empty to list the most recent of everything. Use kind=audit to find audits directly; an account overview is not needed first.

  Parameters:

  • kind — string — Narrow to a kind such as audit, concept, build or project. Leave it out to search everything at once, which is usually what you want.
  • limit — integer — How many to return. 20 by default.
  • query — string — Words to match: a domain, part of a URL, a client name, a caption. Leave it out to list the most recent work instead of searching.
  • since — string — Only things newer than this date, as YYYY-MM-DD.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• fetch read-only

  Free. Reads any single thing by its address, which looks like postclick://audit/<id> and comes from `search`, from `describe`, or from any other tool's answer. A plain id works too if that is all you have. `expand` is what makes this worth calling: pass `*` and you get everything attached in one go, which for an audit means the findings, the buyers and their walkthroughs, the competitor scan, the copy rewrites and the designs. Each arrives with its headline detail; the long write-ups are marked as withheld with the address to fetch them from, because all of them together run to well over a hundred thousand characters and would swamp the answer. Expand first to see the shape, then fetch the two or three that matter. For the audit as finished prose, `get_audit_report` is one call and reads better than any of this. For the audit written up as prose instead, `get_audit_report` is the one you want. For a picture, take the asset_id off a field here and pass it to `get_asset`.

  Parameters:

  • id — string — Source ID returned by search. Alias of uri for ChatGPT retrieval.
  • uri — string — The address, like postclick://audit/<id>. A bare id works too and will be looked up.
  • expand — string — Comma separated names of things to pull in with it, or `*` for all of them. Every answer lists what else is available under available_links, so you can always go deeper without guessing.
  • fields — string — Comma separated field names, when you only need a few. Left out, you get the useful ones.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• get_asset read-only

  Free. Every design, page capture and logo has an asset_id on the field that holds it. Pass it here. Which form you ask for depends entirely on where the picture is going, and getting it wrong is expensive. `as: "url"` is the DEFAULT AND USUALLY RIGHT one: a public link, no login, full size. Use it for anything you publish with publish_page, and for showing somebody a design. `as: "bytes"` returns the picture itself, downscaled to about 1024 across so it fits in a conversation, and you embed it as data:<mime>;base64,<bytes>. Use it ONLY for a document you are writing in a place that refuses to load pictures from other websites, which includes artifacts. At full size one of these designs would be several hundred thousand tokens, so ask for bytes deliberately and not by habit.

  Parameters:

  • asset_id (required) — string — The id from a picture field, like concept:<id>:image.
  • as — string — `url` for a link, which is right for anything you publish and is the default. `bytes` only for a document that cannot load a picture from another website.
  • max_width — integer — How wide the picture comes back when you ask for bytes. 1024 by default, which is readable in a report. Smaller is cheaper.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• publish_page destructive

  Free. Takes HTML you wrote and hosts it at postclick.ppc.io/r/<name>, public, no login, ready to send to a client. A client gets a plain web page, not a link into a chat. We do not template the page, style it or add anything to it: it is exactly the report you wrote. USE PLAIN IMAGE LINKS HERE, from `get_asset` with `as: "url"`. A page we host loads our pictures at full size, so there is no reason to inline them, and a page has a 2 MB ceiling that a single inlined design would eat on its own. Inlining is only for a document you are writing somewhere that blocks pictures from other websites. Publishing the same name again replaces the page and the link keeps working, which is what a report somebody sends every month needs. Use the client's name in the name so they can read it out.

  Parameters:

  • html (required) — string — The whole page with inline CSS. Use image URLs from get_asset; avoid embedding large image data.
  • slug — string — The last part of the address, like acme-october-review. Taken from the title when left out. Pass the same one again to update a report you already sent.
  • title — string — What the report is called.
  • subject_uri — string — The address of the audit or project it was built from, if there is one.

  Hints: read-only no · destructive yes · idempotent yes · open-world yes

• save_note

  Free. Attaches anything you like to any record: who the client is, what they were charged, which findings they have already fixed, when you last sent them a report. Any name, any value. We never look inside it and the audit engine never reads it, so it cannot change what Postclick does. It is there so the work you do around these audits survives between conversations. Saving the same name again replaces it. Read them back with `get_notes`, and they also come back on the record itself.

  Parameters:

  • subject_uri (required) — string — The address of the thing this is about, like postclick://audit/<id>.
  • key (required) — string — Your own name for this field.
  • value — ['string', 'number', 'boolean', 'object', 'array', 'null'] — Anything: a string, a number, a list, an object.

  Hints: read-only no · destructive no · idempotent yes · open-world no

• get_notes read-only

  Free. Returns the notes saved against one record, or every note in the workspace when no address is given. Worth calling at the start of a repeat job, so a monthly report picks up whatever the last one left behind.

  Parameters:

  • subject_uri — string — Limit to one record's notes. Leave it out for all of them.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

Publisher website

Measured by MCPYP probing. Untested fields are omitted, never estimated. Listing is always free; MCPYP never proxies calls or stores user tokens.