Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/Amaculus/screaming-frog-api/llms.txt

Use this file to discover all available pages before exploring further.

Audit helpers are thin, opinionated wrappers that surface the most common SEO issues without requiring you to remember tab names or write manual filters. Each helper returns a list[dict] you can iterate, export, or pass straight into a dataframe.
All helpers work with any crawl backend (DuckDB, Derby, CSV). On lean DuckDB caches, helpers read directly from the prewarmed Derby source so they do not force a full tab materialization first.
Returns broken internal URLs with inlink counts and sampled inlink sources. 
crawl.broken_links_report(
    min_status: int = 400,
    max_status: int = 599,
    max_inlinks: int = 25,
) -> list[dict]
Parameters
Lower bound of the HTTP status code range to flag. Defaults to 400.
Upper bound of the HTTP status code range to flag. Defaults to 599.
Each row includes the broken URL, its HTTP status code, the total inlink count, and up to max_inlinks sampled source URLs. 
from screamingfrog import Crawl

crawl = Crawl.load("./crawl.dbseospider")

broken = crawl.broken_links_report()
for row in broken:
    print(row["Address"], row.get("Status Code"))
To narrow to only client errors, or to widen to all 5xx responses: 
# Only 4xx client errors
client_errors = crawl.broken_links_report(min_status=400, max_status=499)

# Only 5xx server errors, show up to 50 inlink sources per URL
server_errors = crawl.broken_links_report(min_status=500, max_status=599, max_inlinks=50)
The example script in examples/broken_links_report.py shows how to print inlink sources for each broken URL: 
from screamingfrog import Crawl
import sys

crawl = Crawl.load(sys.argv[1] if len(sys.argv) > 1 else "./crawl.dbseospider")

for row in crawl.tab("response_codes_internal_client_error_(4xx)"):
    url = str(row.get("Address") or "")
    code = row.get("Status Code")
    if not url:
        continue
    print(f"{code}: {url}")
    inlinks = list(crawl.inlinks(url))
    for link in inlinks[:25]:
        print(f"  <- {link.source} ({link.anchor_text or ''})")
    if len(inlinks) > 25:
        print(f"  ... {len(inlinks) - 25} more")
Returns the inlinks (source → destination pairs) that point to broken destinations. 
crawl.broken_inlinks_report(
    min_status: int = 400,
    max_status: int = 599,
) -> list[dict]
Parameters
Lower bound of the HTTP status code range. Defaults to 400.
Upper bound of the HTTP status code range. Defaults to 599.
Each row describes an inlink edge — the source page, the broken destination URL, and the HTTP status of the destination. 
from screamingfrog import Crawl

crawl = Crawl.load("./crawl.dbseospider")

for row in crawl.broken_inlinks_report():
    print(row.get("Source"), "->", row.get("Address"), row.get("Status Code"))
Returns all inlinks that carry a nofollow rel attribute. 
crawl.nofollow_inlinks_report() -> list[dict]
Each row includes the source URL, destination URL, anchor text, and the Follow / Rel fields. 
from screamingfrog import Crawl

crawl = Crawl.load("./crawl.dbseospider")

for row in crawl.nofollow_inlinks_report():
    print(row.get("Source"), "->", row.get("Address"), row.get("Rel"))

​
Content helpers

​
title_meta_audit

Surfaces pages with missing titles or missing meta descriptions as flat issue rows. 
crawl.title_meta_audit() -> list[dict]
Each row includes at minimum the Address and an Issue field describing what is missing (Missing Title, Missing Meta Description, etc.).
This helper runs DuckDB-first when internal_all is already cached, and falls back to the high-level internal model on lean caches — so it is fast regardless of cache state.
from screamingfrog import Crawl

crawl = Crawl.load("./crawl.dbseospider")

for row in crawl.title_meta_audit():
    print(row.get("Address"), "|", row.get("Issue"))
The examples/title_meta_audit.py script shows a manual fallback approach that also works against CSV exports: 
from screamingfrog import Crawl
import sys

crawl = Crawl.load(sys.argv[1] if len(sys.argv) > 1 else "./crawl.dbseospider")

# Missing titles — tries the page_titles_missing tab first, falls back to internal scan
print("Missing titles:")
try:
    for row in crawl.tab("page_titles_missing"):
        address = row.get("Address")
        if address:
            print(f"  {address}")
except Exception:
    for page in crawl.internal:
        title = page.data.get("Title 1") or page.data.get("Title")
        if not title:
            print(f"  {page.address}")

# Missing meta descriptions
print("\nMissing meta descriptions:")
try:
    for row in crawl.tab("meta_description_missing"):
        address = row.get("Address")
        if address:
            print(f"  {address}")
except Exception:
    for page in crawl.internal:
        meta = page.data.get("Meta Description 1") or page.data.get("Meta Description")
        if not meta:
            print(f"  {page.address}")

​
Indexability helper

​
indexability_audit

Returns non-indexable pages with the key indexability fields that explain why. 
crawl.indexability_audit() -> list[dict]
Typical fields in each row: Address, Indexability, Indexability Status, Meta Robots 1, X-Robots-Tag 1, Canonical Link Element 1. 
from screamingfrog import Crawl

crawl = Crawl.load("./crawl.dbseospider")

for row in crawl.indexability_audit():
    print(row.get("Address"), "|", row.get("Indexability Status"))
To group non-indexable pages by reason: 
from collections import Counter
from screamingfrog import Crawl

crawl = Crawl.load("./crawl.dbseospider")

counts = Counter(
    row.get("Indexability Status", "Unknown")
    for row in crawl.indexability_audit()
)
for reason, n in counts.most_common():
    print(f"{n:>6}  {reason}")

​
Orphan pages helper

​
orphan_pages_report

Returns pages that have no internal inlinks (i.e., orphaned from the crawl graph). 
crawl.orphan_pages_report(
    ignore_self_links: bool = True,
    only_indexable: bool = False,
) -> list[dict]
Parameters
When True, only indexable orphan pages are returned. Defaults to False (all orphans regardless of indexability).
from screamingfrog import Crawl

crawl = Crawl.load("./crawl.dbseospider")

# All orphan pages
orphans = crawl.orphan_pages_report()

# Indexable orphans only
indexable_orphans = crawl.orphan_pages_report(only_indexable=True)

for row in indexable_orphans:
    print(row.get("Address"))

​
Issue-family helpers

The following helpers map directly to the corresponding issue families in the Screaming Frog UI. Each returns a flat list[dict].
On DuckDB caches, these helpers read issue relations directly when they exist, avoiding redundant tab materialization.

​
security_issues_report

Returns pages with security-related issues (e.g., mixed content, missing HTTPS). 
crawl.security_issues_report() -> list[dict]

from screamingfrog import Crawl

crawl = Crawl.load("./crawl.dbseospider")

for row in crawl.security_issues_report():
    print(row.get("Address"), row.get("Issue"))

​
canonical_issues_report

Returns pages with canonical issues (e.g., non-indexable canonicalised pages, canonical points to redirect). 
crawl.canonical_issues_report() -> list[dict]

from screamingfrog import Crawl

crawl = Crawl.load("./crawl.dbseospider")

for row in crawl.canonical_issues_report():
    print(row.get("Address"), row.get("Canonical Link Element 1"))

​
hreflang_issues_report

Returns pages with hreflang issues (e.g., missing return tags, incorrect language codes). 
crawl.hreflang_issues_report() -> list[dict]
Some hreflang edge cases (specifically incorrect language-code cases) do not yet have exact Derby parity and may differ slightly from the Screaming Frog UI.
from screamingfrog import Crawl

crawl = Crawl.load("./crawl.dbseospider")

for row in crawl.hreflang_issues_report():
    print(row.get("Address"), row.get("Issue"))

​
redirect_issues_report

Returns pages with redirect issues (e.g., redirect chains, redirect loops, broken redirects). 
crawl.redirect_issues_report() -> list[dict]

from screamingfrog import Crawl

crawl = Crawl.load("./crawl.dbseospider")

for row in crawl.redirect_issues_report():
    print(row.get("Address"), row.get("Redirect URL"), row.get("Status Code"))

​
Working with results

All helpers return a plain list[dict], so you can pass them directly to pandas or polars: 
import pandas as pd
from screamingfrog import Crawl

crawl = Crawl.load("./crawl.dbseospider")

df = pd.DataFrame(crawl.broken_links_report())
print(df.groupby("Status Code").size())
You can also use crawl.summary() for a fast top-level count of issues without iterating every row: 
print(crawl.summary())
Issue-family and chain totals in summary() may be None on lean DuckDB caches until those tab families are materialized.

Build docs developers (and LLMs) love

Get started for freeTalk to us