Back to tracker
Plugin tracker
Tools
Explore what tracked Claude Connectors and ChatGPT Plugins can actually do. Search by tool, Plugin, Brand, category, verb, or access requirement.
Latest snapshot2026-09-12USmethodology registry-public-v1
Searchable tools
113,018
Authless tools
7,424
Auth required
100,766
Described tools
61,167
113,018 tools
- Make purchasemake · Validates the agent's intent against Axiom's mandate classifier and starts an end-to-end
agentic checkout in one call, completing the purchase through the merchant's own external checkout.
Field semantics — read carefully, sufficient calls depend on this:
- `userCommand` is the user's original words (audit-grade). Quote near-verbatim. The classifier
uses it to detect intent mismatches against your resolved fields, so don't paraphrase it into
something the user didn't say.
- `merchant.merchantName`, `merchant.merchantURL`, and `itemName` are the AGENT'S RESOLVED
specifics. The user often won't supply these literally — they may say "my dog food" or "the
new iPhone". Your job is to resolve those into a concrete product + merchant when you have the
context to do so faithfully (prior conversation turns, user profile, persistent memory).
- `intentContext` is optional context for terse or elliptical commands. Use it when the literal
userCommand is something like "same one", "try again", "buy that", or "can you get that one?", and your
conversation context clearly points to one concrete purchase. Provide an audit-safe summary and
concrete evidence; do not include hidden chain-of-thought.
- `itemSubtotal` is an optional spend cap the user has explicitly approved; omit when the user
did not state a budget.
Resolution rules (this is what determines whether Axiom can start the purchase):
1. Pass the most SPECIFIC resolved item + merchant you can defend. "Acana Adult Recipe 25lb" beats
"dog food". A merchant URL pointing at the product detail page beats a homepage URL.
2. Pull from your own context — past turns, the user's profile, your memory — to fill specifics
the user implied but didn't state. "My usual" / "again" / possessive references are invitations
for you to resolve from context.
3. Never invent. If the user has not given you, and your context does not contain, the brand /
SKU / size / variant — do NOT make one up. Call the tool with what you have; the
needs_clarification path is the safety net.
4. Required variants for the product type must be resolved: size for clothing/shoes, storage for
phones/laptops, flavor + size for food/drink, loft+grind+finish for golf wedges, etc. If you
can't resolve them faithfully, accept the needs_clarification round-trip.
5. For contextual commands, include `intentContext.evidence` with the specific prior message,
memory, or visible state that supports the resolved purchase. Without that evidence Axiom may
ask for clarification instead of starting checkout.
Responses:
- status='running' with browserPurchaseSessionID — mandate approved, browser session queued.
Call get_purchase_status with that browserPurchaseSessionID to stream progress and retrieve the
receipt, approval link, or clarification request.
- status='needs_clarification' with `question` — the pre-mandate sufficiency check determined
the resolved fields are too ambiguous. Put `question` and any `options` to the user
VERBATIM, then re-call make_purchase with the user's answer folded into the resolved fields
(typically `itemName`, occasionally `merchant.*`).
- status='unavailable' with `reason`, `question`, and `options` — the requested product or
variant was identified but is unavailable. Surface the reason and available options to the user.
- status='mandate_denied' with `reason` — the mandate validator rejected the request (intent
mismatch or prohibited merchant). Surface the reason to the user.
- status='limit_reached' — Axiom's beta purchase limits prevented a new purchase from starting.
Surface the generic message exactly and do not infer or promise a reset time.
- status='account_setup_required' — the user is missing a verified email, linked payment card,
phone number, or complete shipping address. Surface the message and setupUrl exactly. Do not
retry until the user finishes setup in Axiom.
- status='failed', 'cancelled', or 'human_needed' — the browser purchase reached an expected
terminal outcome. Surface the reason to the user; these are not tool-call errors.
Returns IMMEDIATELY in all cases — never waits on the browser pipeline.AxiomAxiom
PluginrequiredData & Analytics - Submit purchase clarificationsubmit · Submits an answer for an awaiting_user_input browser purchase and immediately resumes
the server-side checkout flow on the same browserPurchaseSessionID. Provide either a selectedOptionID,
free-text userInput, or both, then poll get_purchase_status for continued progress.AxiomAxiomPluginrequiredData & Analytics
- App Fetchfetch · Internal APP fetch tool for accessing internal APIsPDF TranslatorSider AIPluginrequiredAI
- pdf Translatorpdf · When a user indicates that they want to translate a file—especially a PDF—call this tool. The user does not need to have uploaded the file beforehand; the tool will guide them through uploading the file after it is invoked.PDF TranslatorSider AIPluginrequiredAI
- add_fieldadd · Append a single field to an existing form.FormspryFormspryPluginrequiredOperations
- close_formclose · Close a form so it stops accepting new responses.FormspryFormspryPluginrequiredOperations
- create_formcreate · Create a new form. Provide a title and, optionally, the fields to include. Returns the shareable form URL.FormspryFormspryPluginrequiredOperations
- get_formget · Get a single form's title, description, status, public URL and its fields.FormspryFormspryPluginrequiredOperations
- get_form_analyticsget · Get a form's stats: views, responses, completion rate, paid responses and revenue (INR).FormspryFormspryPluginrequiredOperations
- get_responsesget · Get recent responses for a form, with answers keyed by field label.FormspryFormspryPluginrequiredOperations
- list_formslist · List all forms on the connected Formspry account, with status, view count and number of responses.FormspryFormspryPluginrequiredOperations
- publish_formpublish · Publish a form so it starts accepting responses.FormspryFormspryPluginrequiredOperations
- update_formupdate · Update a form's title, description, or published state.FormspryFormspryPluginrequiredOperations
- Send a Telegram message or filesendtelegrammessage · Use this tool whenever the user's request or a workflow/task requires sending a Telegram message, image, or supported file (JPEG, PNG, WebP, PDF, TXT, MD, CSV, JSON, or JS), including scheduled or conditional tasks.
When the conditions defined by the user are met, execute this tool without asking for additional confirmation, as that would interrupt the background flow.TaskNotify17labsPluginrequiredProductivity
- Send an email to yourselfsendemail · Use this tool whenever the user's request or a workflow/task requires sending an email to themselves — a summary, a report, a document, or a notification — including scheduled or conditional tasks. The email is always delivered to the address on the user's own TaskNotify account; there is no recipient parameter and this tool cannot email anyone else.
For formatting, use only these HTML tags: <b>, <i>, <u>, <s>, <code>, <pre>, <a href="URL">, <blockquote>, <p>, <br>, <ul>, <ol>, <li>, <h1>–<h3>. Do not use Markdown, CSS, <style>, <div>, <table> or images — they are stripped. Supported attachments: JPEG, PNG, WebP, PDF, TXT, MD, CSV, JSON, up to 12 MB, one per email.
When the conditions defined by the user are met, execute this tool without asking for additional confirmation, as that would interrupt the background flow.TaskNotify17labsPluginrequiredProductivity
- Set up Telegram notificationssetuptelegram · Use this tool when the user wants to set up Telegram notifications for the first time, reconnect, or check whether Telegram is already connected — and before the first call to `sendTelegramMessage` if pairing status is unknown.
For an explicit setup, configure, connect, or switch-chat request, call this tool with `forceReconnect: true` so the user always receives a fresh pairing link, even if this account still has an existing connection. Omit `forceReconnect` only when the user is asking to check the current connection status.
If the result status is "pending_confirmation", show the returned `pairingUrl` to the user as a clickable link and tell them to open it and press "Start" in Telegram. Do not invent your own instructions or link text.
If the user says they already opened the bot without a link (e.g. arrived at the bot directly and it asked them to say "Configure Telegram in TaskNotify"), just call this tool again to get a fresh link.
A stale connection (bot blocked, chat deleted) is detected automatically: if the stored chat is no longer reachable, this returns a fresh "pending_confirmation" link instead of "paired", with wording that explains the old connection broke — show that link the same way as a first-time setup.
If the user explicitly wants to switch to a different Telegram chat while the current one still works, call this again with `forceReconnect: true` to get a new pairing link even though status is already "paired".TaskNotify17labsPluginrequiredProductivity
- Add Recipe To Projectadd · Adds an existing recipe to a project. The project may be in a different material family than the recipe.Uncountable (US)UncountablePluginrequiredData & Analytics
- Archive Recipesarchive · Archives the provided recipes and their child recipes. Fails if any of the recipes are locked. An optional reason can be provided for the archive.Uncountable (US)UncountablePluginrequiredData & Analytics
- Associate Recipe As Inputassociate · Associates a recipe as an input (ingredient). When no existing input is specified, creates a new input or returns the existing association if one already exists. When an existing input is specified, associates it with the recipe; fails if that input is already associated with a different recipe. Use `disassociate_recipe_as_input` to remove an existing association first. Returns the input ID and whether a modification was made.Uncountable (US)UncountablePluginrequiredData & Analytics
- Create Inputscreate · Creates new inputs (ingredients or process parameters) in the specified material families. Each input must belong to at least one material family. Optionally sets attribute values, a category, and field values on creation.Uncountable (US)UncountablePluginrequiredData & Analytics
- Create Recipecreate · Creates a single recipe in the specified material family and workflow. If a matching recipe already exists, no new recipe is created and the existing recipe ID is returned. Returns the recipe ID.Uncountable (US)UncountablePluginrequiredData & Analytics
- Create Recipe Linkcreate · Creates a directional link between two recipes of the specified type (child, control, user_link, or previous_experiment). Skips without error if the link already exists. A name is required for child and user_link types; control and previous_experiment links do not accept a name. Returns whether a modification was made.Uncountable (US)UncountablePluginrequiredData & Analytics
- Disassociate Recipe As Inputdisassociate · Removes the input (ingredient) association from a recipe. If the recipe is not currently associated as an input, this is a no-op. Returns whether a modification was made.Uncountable (US)UncountablePluginrequiredData & Analytics
- Get Current User Infoget · Returns the user ID of the currently authenticated user. Requires a valid authenticated session. Takes no arguments.Uncountable (US)UncountablePluginrequiredData & Analytics
- Get Input Namesget · Returns the ID and name of all non-archived inputs (ingredients and/or process parameters) in a material family. Can filter by type (ingredient vs parameter) using is_parameter. Paginated with a default and maximum limit of 20,000 results, ordered by input ID.Uncountable (US)UncountablePluginrequiredData & Analytics
- Get Inputs Dataget · Returns detailed metadata for a list of inputs by their IDs, including attribute values, categories per material family, and subcategory tags. Also returns attribute definitions (name and quantity type) for all referenced attributes.Uncountable (US)UncountablePluginrequiredData & Analytics
- Get Output Dataget · Returns detailed metadata for outputs in a material family, including quantity type, global category assignments, and associated units. Can filter by specific output IDs. Excludes archived outputs. Paginated with a default and maximum limit of 1,000 results, ordered by output ID. Also returns referenced global category and unit definitions.Uncountable (US)UncountablePluginrequiredData & Analytics
- Get Output Namesget · Returns the ID and name of all non-archived outputs in a material family. Outputs represent measurement values of any type (numeric, text, curve, calculation, etc.). Paginated with a default and maximum limit of 20,000 results, ordered by output ID.Uncountable (US)UncountablePluginrequiredData & Analytics
- Get Projectsget · Lists non-archived projects the user has permission to access, optionally filtered by material family ID. Requires either allMaterialFamilies=true or a specific materialFamilyId; returns 400 if neither is set. Returns projects ordered by ID along with their associated phase definitions. Projects represent groupings of experiments, similar to folders in a directory structure.Uncountable (US)UncountablePluginrequiredData & Analytics
- Get Recipe Linksget · Retrieves links between recipes, traversing up to the specified depth with a maximum depth of 20. Only returns immediate links at each traversal level. Optionally filters by link type. Results are limited to 500 links total; fails if this limit is exceeded.Uncountable (US)UncountablePluginrequiredData & Analytics
- Get Recipe Metadata Dataget · Retrieves recipe metadata field definitions for a material family. Recipe metadata represents data associated with experiments that is not an input, output, or calculation (e.g. sample ID, location). Results are paginated with a default and maximum limit of 1000, ordered by metadata field ID. Optionally filters to specific metadata IDs.Uncountable (US)UncountablePluginrequiredData & Analytics
- Get Recipe Namesget · Returns the ID and name of all non-archived recipes in the specified project, ordered by recipe ID. Archived recipes are excluded from the results.Uncountable (US)UncountablePluginrequiredData & Analytics
- Get Recipes Dataget · Retrieves data for the requested experiments or project, including inputs, outputs, metadata, workflow steps, step relationships, tags, experiment groups, and lock status. Each step includes both `recipeStepNumber` (its stored position within its workflow step) and `displayMixOrderNumber` (the experiment editor's mix-order number). Paginated by experiment ID.Uncountable (US)UncountablePluginrequiredData & Analytics
- List ID Sourcelist · List id and label pairs for a given id source spec. Supports pagination via offset and limit and text search filtering via search_label.Uncountable (US)UncountablePluginrequiredData & Analytics
- Lock Recipeslock · Locks experiments to prevent editing. Supports locking all data (inputs and measurements) or inputs only. Returns whether a modification was made.Uncountable (US)UncountablePluginrequiredData & Analytics
- Remove Recipe From Projectremove · Removes a recipe from a project.Uncountable (US)UncountablePluginrequiredData & Analytics
- Remove Recipe Linkremove · Removes a directional link between two recipes of the specified type (child, control, user_link, or previous_experiment). If the link does not exist, the action is a no-op. Returns whether a modification was made.Uncountable (US)UncountablePluginrequiredData & Analytics
- Set Input Categoryset · Sets or unsets the global category for an input within its material families. Pass a category_id to assign a category, or null to remove the current category. Optionally scope the update to specific material families via material_family_ids; if omitted, updates across all material families the input belongs to. Validates that the category exists. Requires SET_FIELD_VALUE access on the input entity.Uncountable (US)UncountablePluginrequiredData & Analytics
- Set Input Subcategoriesset · Modifies the subcategory (tag) assignments for an input. Supports three update types: 'append' adds the specified subcategories without removing existing ones, 'override' replaces all subcategories with the specified set (archiving any not in the list), and 'remove' archives only the specified subcategories. Requires SET_FIELD_VALUE access on the input entity.Uncountable (US)UncountablePluginrequiredData & Analytics
- Set Recipe Inputsset · Sets input values for experiments. Supports numeric, text, and categorical types. Validates that the recipe step belongs to the correct recipe. Supports setting actual values, associating lots (creating them if needed), and removing inputs. Writes all values to the first workflow step when no step is specified. Fails if both a numeric and string value are provided.Uncountable (US)UncountablePluginrequiredData & Analytics
- Set Recipe Metadataset · Sets metadata values on a recipe identified by key. Accepts a list of metadata field-value pairs to populate on the recipe. Fails if any of the metadata fields are read-only.Uncountable (US)UncountablePluginrequiredData & Analytics
- Set Recipe Outputsset · Sets output measurement values for experiments. Supports numeric, text, categorical, curve, color, and recipe-as-output types. Fails if both a numeric and string value are provided. Experiment number must be 1 or greater. Use `resolve_output_conditions` to obtain a condition ID when setting conditioned outputs. Supports configurable null behavior (do nothing, ensure output exists, or remove the output row).Uncountable (US)UncountablePluginrequiredData & Analytics
- Set Recipe Tagsset · Modifies tags on a recipe identified by key. Supports three update modes - append (adds tags to existing ones), override (replaces all tags with the provided set), or remove (removes the specified tags).Uncountable (US)UncountablePluginrequiredData & Analytics
- Unarchive Recipesunarchive · Restores previously archived recipes, making them active again. Fails if any of the recipes are locked.Uncountable (US)UncountablePluginrequiredData & Analytics
- Unlock Recipesunlock · Unlocks previously locked experiments, making them editable again. Returns whether a modification was made.Uncountable (US)UncountablePluginrequiredData & Analytics
- 교보문고 도서 추천getbestsellers · One of three book-recommendation tools, in priority order (getKeywordBookList / searchBooks / getBestsellers). Use THIS tool for broad, general bestseller requests: either an unqualified "recommend me a book" with no specific angle, or a request that clearly matches one of the top-level categories below. This is also the default fallback whenever the request is too vague for the other two tools.
When to use this tool:
- "recommend a good book", "아무거나 추천해줘", "read anything good lately?" → TOT (all)
- "popular novels these days" → 01 (Fiction)
- "trending business books" → 13 (Business)
- "best self-help books" → 15 (Self-help)
- "popular books for kids" → 42 (Children)
Use getKeywordBookList instead when the request reduces to a single mood/season/genre/occasion keyword (e.g. "여름", "공포", "힐링", "추리", "크리스마스", "설렘", "인생책") — it queries curated editorial tags and returns more relevant picks than a category ranking would.
Use searchBooks instead for a specific book title/author, or a descriptive request that needs more than one concept and doesn't map to a category or a single keyword (e.g. "이직 준비할 때 읽으면 좋은 책", "부모님 선물용 자기계발서").
Category codes (top-level only):
TOT=All, 01=Fiction, 03=Poetry/Essay, 05=Humanities, 07=Family/Parenting, 08=Cooking, 09=Health, 11=Hobby/Sports, 13=Business/Economics, 15=Self-help, 17=Politics/Society, 19=History/Culture, 21=Religion, 23=Arts/Pop-culture, 25=Middle-High-school, 26=Engineering, 27=Foreign-language, 29=Science, 31=Exam-prep, 32=Travel, 33=IT/Computer, 35=Magazine, 38=Youth, 39=Elementary, 41=Infant(0-7), 42=Children, 47=Comics, 50=University, 53=Korea-intro, 59=Kyobo-original.
Rules:
- If user intent matches a top-level category above → use this tool with that code
- If uncertain, too broad, or no specific angle at all → default to TOT
- If user mentions a specific date/period → set ymw param
ymw format: monthly YYYYMM (e.g. 202607), weekly/daily YYYYMMDD (e.g. 20260727)
IMPORTANT — daily (period=001) ranking availability: the API has NOT finished aggregating the current day's ranking yet, so setting ymw to TODAY'S date always returns an empty result. "오늘/today's bestseller"-style requests with no explicit date do NOT mean "set ymw to today" — omit ymw entirely instead, which returns the latest finalized ranking (normally the previous day's). Only set ymw explicitly to a past date (YYYYMMDD) when the user names a specific day (e.g. "어제", "지난 화요일", or an explicit date) — and never to today's date.교보문고Kyobo Book CentrePluginnoneCommerce
- 교보문고 매장 상세 조회getstoredetail · Shows one Kyobo Book Centre offline store's location, business hours, and contact info on a map. Use THIS tool only when the user's question names a specific store/branch (e.g. "광화문점 영업시간 알려줘", "강남점 전화번호가 뭐야"). For a book search scoped to a specific store, use searchStoreBooks instead; for general book recommendation/search, use getBestsellers/searchBooks/getKeywordBookList instead.
Either storeName or storeCode must be provided (storeName is a LIKE search, so a partial name like "강남" or "광화문" is enough to match).
Examples:
- "광화문점 영업시간 알려줘" → storeName: "광화문"교보문고Kyobo Book CentrePluginnoneCommerce
- 도서 검색searchbooks · One of three book-recommendation tools, in priority order (getKeywordBookList / searchBooks / getBestsellers). Use THIS tool when:
1. The user mentions a specific book title or author name
2. The user has a descriptive reading need that combines more than one concept and doesn't reduce to a single mood/genre/season keyword and doesn't match a bestseller top-level category
Examples for case 1 (specific title/author):
- "Find Demian" → keyword: "데미안"
- "Search for Harry Potter" → keyword: "해리포터"
Examples for case 2 (multi-concept descriptive need → extract keyword):
- "이직 준비할 때 읽으면 좋은 책" → keyword: "이직"
- "돈 관리 잘하는 법 알려주는 책" → keyword: "재테크"
- "면접 잘 보는 법 알려주는 책" → keyword: "면접"
- "부모님 선물용 자기계발서" → keyword: "자기계발"
- "Books about AI ethics" → keyword: "인공지능윤리"
Rules:
- Extract the most relevant Korean keyword from user intent
- For a single mood/season/genre/occasion keyword (e.g. "여름", "공포", "추리", "힐링", "크리스마스") → use getKeywordBookList instead — it queries curated editorial tags and returns more relevant picks
- For broad top-level categories, or an unqualified "recommend a book" (fiction, self-help, business, etc.) → use getBestsellers instead교보문고Kyobo Book CentrePluginnoneCommerce
- 매장 도서 검색searchstorebooks · Finds books in stock at a specific Kyobo Book Centre offline store, by search term. Use THIS tool when a store name and a search term are mentioned together, e.g. "광화문점에 이 책 있어?", "강남점에서 자기계발서 찾아줘". If the request is only about the store itself (business hours/phone number), not book stock, use getStoreDetail instead.
keyword works with genre/topic words just as well as exact book titles — if the request combines a recommendation with a specific store (e.g. "광화문점에 괜찮은 인문 서적 있어?"), don't chain a separate book-recommendation tool first; call this tool directly with keyword: "인문".
storeName must match one of the stores listed below (a partial name is fine, e.g. "광화문" → "광화문점"). If the store isn't in the list, don't call this tool — tell the user instead.
Store list: 광화문점, 가든파이브점, 강남점, 건대스타시티점, 동대문점, 목동점, 서울대점, 수유점, 영등포점, 원그로브점, 은평점, 이화여대점, 잠실점, 천호점, 청량리점, 합정점, 광교점, 분당점, 송도점, 인천점, 일산점, 판교점, 평촌점, 수원점, 경성대ㆍ부경대점, 광주상무점, 대구점, 대전점, 부산점, 세종점, 센텀시티점, 울산점, 전주점, 창원점, 천안점, 칠곡점, 해운대 팝업스토어교보문고Kyobo Book CentrePluginnoneCommerce
- 인물 검색searchauthors · STEP 1 of the two-step person lookup (searchAuthors → getAuthorDetail). This is a LIST search: it returns every person (author, translator, illustrator) at Kyobo Book Centre whose name matches, with a short bio and representative works for each. It does NOT return full profile details — getAuthorDetail does.
Use THIS tool when the user asks ABOUT a person:
- "한강 작가에 대해 알려줘" → searchText: "한강"
- "베르나르 베르베르가 누구야?" → searchText: "베르나르 베르베르"
- "김영하 작가 프로필" → searchText: "김영하"
- "Who is Haruki Murakami?" → searchText: "무라카미 하루키"
- "이 작가 어떤 사람이야?" (a specific name is in context) → searchText: that name
Use searchBooks instead when the user wants that person's BOOKS rather than their profile:
- "한강 책 추천해줘" / "김영하 소설 뭐 있어?" → searchBooks
MANDATORY follow-up — always confirm before the detail lookup:
The API matches names by SUBSTRING, so a search for "한강" also returns 곽한강, 심한강, 이한강 — people the user never asked about. Never assume the first entry is the right one.
1. Present the returned people (name + what distinguishes them, e.g. their representative work)
2. ASK the user which one they meant — e.g. "말씀하신 분이 『소년이 온다』를 쓴 한강 작가가 맞을까요?" If several plausibly match, list them and ask which one.
3. WAIT for the user's answer. Do NOT call getAuthorDetail in the same turn as this search.
4. Once the user confirms, call getAuthorDetail with that person's chrcCode from this result.
The only exception: if exactly one person is returned AND their name matches exactly, you may still show them and confirm — but never skip straight to getAuthorDetail on an ambiguous match.
Rules:
- searchText must be a person's name only — never a book title, genre, or theme
- Korean names work best; romanized names may not match
- Present only the entries whose name actually matches what the user asked about교보문고Kyobo Book CentrePluginnoneCommerce
What is Tool Explorer?
Tool Explorer indexes the callable tool names and descriptions attached to public registry profiles. It is useful for seeing what agents can actually invoke, not just which profile exists.
How do category and verb filters work?
Category filters use the live registry category rollup. Verb filters use the public tool insights rollup, so the page stays backed by the same read models as the tracker charts.
Why do auth requirements matter?
Auth requirements show whether a tool is likely usable without account connection, requires authentication, is private, or is unknown in the current snapshot.