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 snapshot2026-09-11USmethodology registry-public-v1
Searchable tools
113,017
Authless tools
7,424
Auth required
100,766
Described tools
61,166
113,017 tools
- Send replysend · Reply in an existing thread from the customer's mailbox. Two-step: call without confirm to get the rendered reply (from, to, subject, body) and an approval_token; show it to the user; then call again with the unchanged body, the approval_token, and confirm: true. Exact retries never send twice.DayOneLeadDayOneLead
PluginrequiredSales & CRM - Set CRM statusset · Add a lead to the CRM or move it: new, working, followup, won, lost. Moving off 'new' stamps last_contacted_at unless you pass one.DayOneLeadDayOneLeadPluginrequiredSales & CRM
- Update matching profileupdate · Update the ideal customer profile (owner only). Pass only the fields to change. Next morning's picks use it.DayOneLeadDayOneLeadPluginrequiredSales & CRM
- pause campaignpause · Use this when the user wants to pause or temporarily stop an active email campaign. Queued sends are cancelled and resume where they left off.DayOneLeadDayOneLeadPluginrequiredSales & CRM
- resume campaignresume · Use this when the user wants to resume or restart a paused email campaign.DayOneLeadDayOneLeadPluginrequiredSales & CRM
- Get a businessget · Full detail for one business: live status, how long it stays open, its regular weekly hours, and how fresh the information is. Identify it by `business_id` (UUID) or `slug`.OPEN - what is open right nowOPENPluginnoneConsumer & Lifestyle
- List categorieslist · The available business categories and their slugs. Call this before passing `category` to the other tools.OPEN - what is open right nowOPENPluginnoneConsumer & Lifestyle
- Search businessessearch · Search OPEN businesses by name or description, with each one's live status. Use this when the user names a place or a kind of place; use `open_now` when they ask what is open around them. Renders an interactive map.OPEN - what is open right nowOPENPluginnoneConsumer & Lifestyle
- What is open nowopen · Find businesses that are open RIGHT NOW near a location, with how long they stay open. Pass latitude+longitude when known, otherwise pass `place` as free text (e.g. "Machane Yehuda", "מחנה יהודה") and it will be geocoded. With neither, results are centred on central Jerusalem and the result says so explicitly - tell the user it is a default rather than implying you know where they are. Renders an interactive map.OPEN - what is open right nowOPENPluginnoneConsumer & Lifestyle
- Hämta en publicerad promptmallget · Fetch one template in full, ready to use, by the id returned from search_templates or list_templates. context_keys only chooses which stored variant comes back. The template arrives as its own text plus the fields the user is meant to fill in; the client fills those in locally -- the server never returns a finished, filled-in prompt.PromptbankenPromptbankenPluginnoneAI
- Hämta ett publicerat promptpaketget · Fetch one package by its slug -- a short name such as 'kommunikation', not a template id -- in the variant that best matches context_keys. Returns the package's introduction plus the fields the client fills in locally; the server never returns finished text. Use list_package_prompts for the prompts inside the package.PromptbankenPromptbankenPluginnoneAI
- Hämta routing- och integritetsreglerget · Get Promptbanken's short bootstrap for using this connector: which tool to reach for in which situation, and the privacy rule that the user's own text, personal data and confidential material stay in the client and are never sent to the server. Worth calling once before the first catalogue lookup in a conversation.PromptbankenPromptbankenPluginnoneAI
- Kontrollera tjänstens statushealth · Diagnostic only: check that Promptbanken is reachable and report the service status. Not part of normal use -- go straight to search_templates or recommend_packages for a user's task, and reach for this only when a call has failed and you need to know whether the service is up. Reads no prompt content and takes no input.PromptbankenPromptbankenPluginnoneAI
- Lista mallar i ett promptpaketlist · List the steps inside one package, in the order they are meant to be used. Takes the package slug -- a short name such as 'kommunikation', not a template id -- and context_keys to pick the variant written for the user's setting. Returns each step's id, title and purpose; fetch a step's prompt text with get_template(id) when that step is actually reached. Set include_prompt_text=true only to pull every step's text at once.PromptbankenPromptbankenPluginnoneAI
- Lista publicerade promptmallarlist · Browse Promptbanken's catalogue of ready-made prompt templates for Swedish public-sector work -- writing, communication, decision support, leadership and more. The catalogue is open to everyone; no account or key is needed. Use search_templates when the task is known -- this tool is for browsing, and is paginated: it returns 25 summaries at a time, with total and has_more telling you whether to fetch the next page with offset. Call get_template(id) for the full text of a chosen template. Set include_prompt_text=true only when every prompt really is needed at once -- it will crowd out the conversation. context_keys picks the variant written for the user's setting, for example kommun or skola.PromptbankenPromptbankenPluginnoneAI
- Lista publicerade promptpaketlist · List Promptbanken's packages -- themed sets of prompts that carry a whole task from start to finish, such as change communication or preparing a decision. Returns just enough to choose one: slug, title, summary and type. Call get_package(slug) for the package's own introduction, and list_package_prompts(slug) for its steps. context_keys picks the variant written for the user's setting; package_type narrows the list when supplied.PromptbankenPromptbankenPluginnoneAI
- Rekommendera promptpaket för en rollrecommend · Suggest which prompt packages suit a given job role -- a good first step when the user does not yet know what to ask for. Takes a short Swedish role term such as 'chef' or 'kommunikator'; ask the user for their role first if it is unknown. An unrecognised role returns every package with role_recognized=false rather than an empty result.PromptbankenPromptbankenPluginnoneAI
- Sök publicerade promptmallarsearch · Find the prompt templates that fit a task, without pulling in the whole catalogue. This is the normal way into the catalogue. PRIVACY: send only general, anonymised search terms that describe the kind of task -- never the user's own text, pasted document content, case material, personal data or confidential information. Summarise the need in a few words instead. The free-text query is matched against title, purpose, tags and output format; area and risk_level narrow the result. role ranks results toward a job function -- it does not hide templates from other areas. context_keys picks the variant written for the user's setting. Returns short summaries, so call get_template(id) on a chosen result for the full prompt.PromptbankenPromptbankenPluginnoneAI
- Create a verified PE practice setcreate · Use this when the user wants a short reviewed PE drill. After calling it, present every returned question and all A–D choices in the chat, then ask the user to choose an answer. Do not reveal the answer until it is graded.PE Prep Pro EngineeringPE Prep ProPluginnoneEducation
- Get a reviewed PE concept briefget · Use this when the user wants a compact reviewed concept summary and primary reference starting points for a concept ID from list_disciplines.PE Prep Pro EngineeringPE Prep ProPluginnoneEducation
- Grade a verified PE practice setgrade · Use this after create_practice_set when the user has answered every question; it grades server-side and returns reviewed explanations, sources, and revision metadata.PE Prep Pro EngineeringPE Prep ProPluginnoneEducation
- List PE disciplines and reviewed topicslist · Use this when the user wants to see available PE disciplines, reviewed topics, counts, concept IDs, or exam-spec metadata.PE Prep Pro EngineeringPE Prep ProPluginnoneEducation
- Report a PE Prep Pro content issuereport · Use this when the user wants to report a specific content issue for editorial review. Do not include personal or confidential information.PE Prep Pro EngineeringPE Prep ProPluginnoneEducation
- cancel_agent_approvalcancel · Cancel one pending approval request owned by this agent identity.JoinLayerJoinLayerPluginrequiredData & Analytics
- cancel_connection_setupcancel · Cancel an unused secure connection setup link.JoinLayerJoinLayerPluginrequiredData & Analytics
- create_connection_setupcreate · Create a browser setup from non-secret config_schema fields. Omit every secret-marked field, credential, and secret placeholder.JoinLayerJoinLayerPluginrequiredData & Analytics
- create_pipeline_draftcreate · Create a pipeline draft. Inspect connections first; validate and preview before any run.JoinLayerJoinLayerPluginrequiredData & Analytics
- diagnose_run_failurediagnose · Return run state, structured errors, events, and dead-letter summaries for diagnosis.JoinLayerJoinLayerPluginrequiredData & Analytics
- discover_connection_schemadiscover · Discover relational schemas, ClickHouse databases, BigQuery datasets and tables, or sample Kafka fields with topic as table and no schema.JoinLayerJoinLayerPluginrequiredData & Analytics
- get_connectionget · Get one connection's safe metadata and current configuration capabilities.JoinLayerJoinLayerPluginrequiredData & Analytics
- get_connection_setup_statusget · Poll a secure connection setup session; completed responses expose only the durable connection ID.JoinLayerJoinLayerPluginrequiredData & Analytics
- get_pipelineget · Get the complete safe pipeline definition, mappings, transforms, checks, setup state, and runtime summary.JoinLayerJoinLayerPluginrequiredData & Analytics
- get_runget · Get run progress, counters, worker lease, recovery state, and curated failure details.JoinLayerJoinLayerPluginrequiredData & Analytics
- get_usage_reportget · Return workspace and per-pipeline usage attribution for an ISO-8601 date range.JoinLayerJoinLayerPluginrequiredData & Analytics
- get_workspace_capacityget · Return current workspace plan capacity, usage, remaining limits, and run blockers.JoinLayerJoinLayerPluginrequiredData & Analytics
- get_workspace_contextget · Return identity, workspace, role, scopes, and whether workspace policy requires approvals for agent run starts/stops.JoinLayerJoinLayerPluginrequiredData & Analytics
- get_workspace_overviewget · Inspect identity, role-visible capacity, connections, and pipelines in one read-only first-session operation.JoinLayerJoinLayerPluginrequiredData & Analytics
- list_activitylist · List workspace run activity using explicit status, mode, period, and pipeline filters.JoinLayerJoinLayerPluginrequiredData & Analytics
- list_agent_approvalslist · List this agent identity's pending and historical approval requests.JoinLayerJoinLayerPluginrequiredData & Analytics
- list_connection_setupslist · List this identity's recent setup sessions to recover safely after an uncertain create response.JoinLayerJoinLayerPluginrequiredData & Analytics
- list_connectionslist · List workspace connections in a paginated object without stored credentials or encrypted configuration.JoinLayerJoinLayerPluginrequiredData & Analytics
- list_connector_typeslist · List providers, roles, capabilities, and exact config_schema fields/defaults/options/secret flags. Call before setup.JoinLayerJoinLayerPluginrequiredData & Analytics
- list_pipeline_runslist · List recent runs and recovery state for a pipeline.JoinLayerJoinLayerPluginrequiredData & Analytics
- list_pipelineslist · List pipelines with one authoritative inventory_state, plus setup, runtime, validation, and matching state summaries.JoinLayerJoinLayerPluginrequiredData & Analytics
- preview_pipelinepreview · Preview transformed and enriched target rows without starting the pipeline.JoinLayerJoinLayerPluginrequiredData & Analytics
- request_run_start_approvalrequest · Request exact human approval when workspace policy requires it. If approval_required=false, start without approval_id; otherwise present approval_url and poll.JoinLayerJoinLayerPluginrequiredData & Analytics
- request_run_stop_approvalrequest · Request human approval to stop one exact run when workspace policy requires it. Honor approval_required and return approval_url/next_action when required.JoinLayerJoinLayerPluginrequiredData & Analytics
- start_pipelinestart · Start or resume a validated pipeline using an approved request or matching automation policy.JoinLayerJoinLayerPluginrequiredData & Analytics
- stop_runstop · Stop an active run using an approved request or matching automation policy.JoinLayerJoinLayerPluginrequiredData & Analytics
- test_connectiontest · Test one stored connection using server-side secrets. Kafka verifies broker/topic/Registry; ClickHouse verifies transport, credentials, and database access.JoinLayerJoinLayerPluginrequiredData & Analytics
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.