Tool catalog

The complete set of tools exposed via the MCP gateway. Auto-synthesized from the REST OpenAPI spec’s MCP tag — see the machine-readable list for full input/output schemas.

An interactive Redoc-style per-tool browser is on the roadmap. Until then, the categories below + the live OpenAPI spec are the source of truth.

API key management

Tool	What it does	Scope required
`scrapewise_whoami`	Identify the caller (returns customerRef + scope + key prefix)	any
`scrapewise_generate_api_key`	Mint a new API key	`LLM_FULL` (USER for the REST equivalent)
`scrapewise_list_api_keys`	List your keys	`LLM_READ`
`scrapewise_delete_api_key`	Revoke a key	`LLM_FULL`

Scrapers — CRUD

Tool	What it does	Scope
`scrapewise_list_scrapers`	List all scrapers for your customer	`LLM_READ`
`scrapewise_get_scraper`	Get one scraper by id	`LLM_READ`
`scrapewise_create_scraper`	Create a new scraper definition	`LLM_FULL`
`scrapewise_update_scraper`	Modify scraper (full PUT)	`LLM_FULL`
`scrapewise_delete_scraper`	Delete scraper (data preserved separately)	`LLM_FULL`

Scrapers — runtime

Tool	What it does	Scope
`scrapewise_run_scraper`	Trigger an async job	`LLM_FULL`
`scrapewise_cancel_scraper_job`	Gracefully cancel in-flight job	`LLM_FULL`
`scrapewise_stop_scraper_job`	Hard-stop in-flight job	`LLM_FULL`
`scrapewise_stream_job_status`	SSE stream of job progress	`LLM_READ`

Scraper groups

Tool	What it does	Scope
`scrapewise_list_scraper_groups`	List groups	`LLM_READ`
`scrapewise_create_scraper_group`	Create a group	`LLM_FULL`
`scrapewise_update_scraper_group`	Rename / re-parent a group	`LLM_FULL`
`scrapewise_delete_scraper_group`	Delete group (scrapers inside become top-level)	`LLM_FULL`
`scrapewise_run_group`	Run all scrapers in a group	`LLM_FULL`

Scraped data

Tool	What it does	Scope
`scrapewise_list_scraper_data`	Paginated read of rows	`LLM_READ`
`scrapewise_delete_scraper_data`	Delete rows (by scraper / group / job / id)	`LLM_FULL`
`scrapewise_export_scraper_data`	Export rows as xlsx / csv / jsonl	`LLM_READ`

Preview

Tool	What it does	Scope
`scrapewise_preview_scraper_from_url`	One-shot scrape, no scraper saved	`LLM_READ`
`scrapewise_preview_scraper_from_curl`	Same but with a curl command (cookies, custom POST, etc.)	`LLM_READ`

Schemas

Tool	What it does	Scope
`scrapewise_list_schemas`	List reusable customer schemas	`LLM_READ`
`scrapewise_get_schema`	Fetch one schema	`LLM_READ`
`scrapewise_create_schema`	Define a new reusable schema	`LLM_FULL`
`scrapewise_update_schema`	Modify a schema (PUT)	`LLM_FULL`
`scrapewise_delete_schema`	Delete a schema (scrapers using it keep their inline copy)	`LLM_FULL`

Sitemaps

Tool	What it does	Scope
`scrapewise_list_sitemap_urls`	List URLs known to be on a site	`LLM_READ`
`scrapewise_parse_sitemap`	Parse a sitemap.xml URL	`LLM_READ`

What’s NOT exposed via MCP

Some REST endpoints are deliberately excluded from MCP:

Endpoint	Why
`/api/admin/*`	Root-auth only
`/api/customer/*`	Customer/account management — too sensitive
`/api/llm/*`	Internal LLM endpoints
`/api/matcher/*`	Internal matcher service
`/api/portal/`, `/api/portal-events/`	Portal UI specific
`/api/plan-features`, `/api/customer-switch/*`	Billing / plan admin

If you need to call one of these, use the REST surface with a USER-scope key (see REST API overview).

Annotations

Each tool carries machine-readable annotations the agent can reason about:

Annotation	Meaning
`readOnly: true`	Tool does not mutate state
`destructive: true`	Tool can delete data (irreversible)
`idempotent: true`	Safe to retry; same inputs → same effect
`openWorld: true`	Tool’s side effects extend beyond Scrapewise (e.g. `run_scraper` makes outbound HTTP requests to target sites)

Well-behaved clients (Claude Desktop, claude.ai) prompt for confirmation on destructive tools by default.

What’s next

Wire Claude to use these tools → MCP quickstart
Prompt-injection contract → Prompt injection
Underlying REST surface → REST API overview