lfx-enconvert 0.0.1__tar.gz

lfx_enconvert-0.0.1/CHANGELOG.md

@@ -0,0 +1,20 @@
+# Changelog
+
+All notable changes to `lfx-enconvert` are documented here.
+The format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [0.0.1] - 2026-08-27
+
+### Added
+
+- First release: the **EnConvert** bundle for Langflow with six components.
+  - **Perceive URL** — `POST /v2/perceive`, with a `render_quality` score on every read.
+  - **Web Search** — `POST /v2/search`.
+  - **Discover URLs** — `POST /v2/discover`.
+  - **Extract Structured** — `POST /v2/distill`.
+  - **Convert File to Markdown** — `POST /v1/convert/anything-to-markdown` (fetches the file URL, then converts).
+  - **Convert File to PDF** — `POST /v1/convert/anything-to-pdf` (fetches the file URL, then converts).
+- Pip-installable extension bundle, auto-discovered by Langflow at server start via the
+  `langflow.extensions` entry point.
+- Private-key (`sk_`) auth over the `X-API-Key` header; public `pk_` keys are rejected with a clear message.

lfx_enconvert-0.0.1/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 EnConvert
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

lfx_enconvert-0.0.1/PKG-INFO

@@ -0,0 +1,57 @@
+Metadata-Version: 2.5
+Name: lfx-enconvert
+Version: 0.0.1
+Summary: EnConvert bundle for Langflow: perceive URLs, search the web, discover URLs, extract structured data, and convert files — every read scored with render_quality.
+Project-URL: Homepage, https://www.enconvert.com
+Project-URL: Documentation, https://www.enconvert.com/docs
+Project-URL: Repository, https://github.com/enconvert/lfx-enconvert
+Author-email: EnConvert <support@enconvert.com>
+License-Expression: MIT
+License-File: LICENSE
+Keywords: bundle,components,enconvert,langflow,lfx,markdown,rag,web
+Requires-Python: >=3.10
+Requires-Dist: lfx<2.0.0
+Requires-Dist: requests>=2.28
+Description-Content-Type: text/markdown
+
+# EnConvert bundle for Langflow
+
+`lfx-enconvert` adds six [EnConvert](https://www.enconvert.com) components to Langflow.
+EnConvert renders web pages and files into agent-ready data, and puts a `render_quality`
+honesty score (0.0–1.0) on every read — so a blocked or empty page comes back flagged, not trusted.
+
+## Components
+
+| Component | Endpoint | What it does |
+|-----------|----------|--------------|
+| **Perceive URL** | `POST /v2/perceive` | A page → markdown, structured data, screenshot, PDF, links… with a `render_quality` score |
+| **Web Search** | `POST /v2/search` | Web / news / images / scholar / patents / maps results |
+| **Discover URLs** | `POST /v2/discover` | Map a site (sitemap, crawl, hybrid) into a list of URLs |
+| **Extract Structured** | `POST /v2/distill` | Typed data from pages against a JSON schema |
+| **Convert File to Markdown** | `POST /v1/convert/anything-to-markdown` | Any file URL → clean markdown |
+| **Convert File to PDF** | `POST /v1/convert/anything-to-pdf` | Any file URL → PDF |
+
+## Install
+
+```bash
+pip install lfx-enconvert
+```
+
+Langflow auto-discovers the bundle at server start. The six components appear under the
+**EnConvert** bundle in the component sidebar — no config, no manual registration.
+
+## API key
+
+Each component has an **EnConvert API Key** field. Use a **private** key (`sk_…`) from your
+[dashboard](https://www.enconvert.com/dashboard/api-keys); public `pk_` keys are rejected.
+Store it once as a Langflow **global variable** and select it in the key field, so it never
+lives in the flow JSON.
+
+## Publishing / distribution
+
+Building, publishing to PyPI, the single-file drop-in route, and the optional core PR are all in
+the **[deploy guide](../langflow-enconvert-deploy/README.md)**.
+
+## Licence
+
+[MIT](LICENSE)

lfx_enconvert-0.0.1/README.md

@@ -0,0 +1,41 @@
+# EnConvert bundle for Langflow
+
+`lfx-enconvert` adds six [EnConvert](https://www.enconvert.com) components to Langflow.
+EnConvert renders web pages and files into agent-ready data, and puts a `render_quality`
+honesty score (0.0–1.0) on every read — so a blocked or empty page comes back flagged, not trusted.
+
+## Components
+
+| Component | Endpoint | What it does |
+|-----------|----------|--------------|
+| **Perceive URL** | `POST /v2/perceive` | A page → markdown, structured data, screenshot, PDF, links… with a `render_quality` score |
+| **Web Search** | `POST /v2/search` | Web / news / images / scholar / patents / maps results |
+| **Discover URLs** | `POST /v2/discover` | Map a site (sitemap, crawl, hybrid) into a list of URLs |
+| **Extract Structured** | `POST /v2/distill` | Typed data from pages against a JSON schema |
+| **Convert File to Markdown** | `POST /v1/convert/anything-to-markdown` | Any file URL → clean markdown |
+| **Convert File to PDF** | `POST /v1/convert/anything-to-pdf` | Any file URL → PDF |
+
+## Install
+
+```bash
+pip install lfx-enconvert
+```
+
+Langflow auto-discovers the bundle at server start. The six components appear under the
+**EnConvert** bundle in the component sidebar — no config, no manual registration.
+
+## API key
+
+Each component has an **EnConvert API Key** field. Use a **private** key (`sk_…`) from your
+[dashboard](https://www.enconvert.com/dashboard/api-keys); public `pk_` keys are rejected.
+Store it once as a Langflow **global variable** and select it in the key field, so it never
+lives in the flow JSON.
+
+## Publishing / distribution
+
+Building, publishing to PyPI, the single-file drop-in route, and the optional core PR are all in
+the **[deploy guide](../langflow-enconvert-deploy/README.md)**.
+
+## Licence
+
+[MIT](LICENSE)

lfx_enconvert-0.0.1/pyproject.toml

@@ -0,0 +1,35 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "lfx-enconvert"
+version = "0.0.1"
+description = "EnConvert bundle for Langflow: perceive URLs, search the web, discover URLs, extract structured data, and convert files — every read scored with render_quality."
+readme = "README.md"
+license = "MIT"
+requires-python = ">=3.10"
+authors = [{ name = "EnConvert", email = "support@enconvert.com" }]
+keywords = ["langflow", "lfx", "enconvert", "bundle", "components", "rag", "web", "markdown"]
+dependencies = [
+    "lfx<2.0.0",
+    "requests>=2.28",
+]
+
+[project.urls]
+Homepage = "https://www.enconvert.com"
+Documentation = "https://www.enconvert.com/docs"
+Repository = "https://github.com/enconvert/lfx-enconvert"
+
+# Langflow discovers installed extension bundles at server start through this entry point.
+# Editable installs whose dist.files only surface dist-info entries fall back to this
+# entry point to locate the extension.json manifest.
+[project.entry-points."langflow.extensions"]
+lfx-enconvert = "lfx_enconvert"
+
+[tool.hatch.build.targets.wheel]
+# extension.json + the component files must live inside the lfx_enconvert package so
+# importlib.metadata.files(dist) finds the manifest and the loader resolves
+# bundles[].path relative to it. The explicit include guarantees the JSON ships.
+packages = ["src/lfx_enconvert"]
+include = ["src/lfx_enconvert/extension.json", "src/lfx_enconvert/components/**/*.py"]

lfx_enconvert-0.0.1/src/lfx_enconvert/__init__.py

@@ -0,0 +1,7 @@
+"""lfx-enconvert — the EnConvert extension bundle for Langflow.
+
+Langflow discovers this package at server start via the ``langflow.extensions``
+entry point (see pyproject.toml) and reads the co-located ``extension.json`` manifest.
+"""
+
+__version__ = "0.0.1"

lfx_enconvert-0.0.1/src/lfx_enconvert/components/__init__.py

@@ -0,0 +1 @@
+"""Component bundles for the lfx-enconvert extension."""

lfx_enconvert-0.0.1/src/lfx_enconvert/components/enconvert/__init__.py

@@ -0,0 +1,17 @@
+"""EnConvert bundle — six components that turn web pages and files into agent-ready data."""
+
+from .convert_markdown import EnConvertConvertToMarkdown
+from .convert_pdf import EnConvertConvertToPdf
+from .discover_urls import EnConvertDiscoverUrls
+from .extract_structured import EnConvertExtractStructured
+from .perceive import EnConvertPerceive
+from .web_search import EnConvertWebSearch
+
+__all__ = [
+    "EnConvertPerceive",
+    "EnConvertWebSearch",
+    "EnConvertDiscoverUrls",
+    "EnConvertExtractStructured",
+    "EnConvertConvertToMarkdown",
+    "EnConvertConvertToPdf",
+]

lfx_enconvert-0.0.1/src/lfx_enconvert/components/enconvert/convert_markdown.py

@@ -0,0 +1,78 @@
+"""Convert File to Markdown — any file URL into clean markdown."""
+
+from __future__ import annotations
+
+import io
+
+import requests
+from lfx.custom.custom_component.component import Component
+from lfx.io import MessageTextInput, Output, SecretStrInput
+from lfx.schema.data import Data
+
+BASE_URL = "https://api.enconvert.com"
+TIMEOUT = 120
+
+
+def _convert(endpoint: str, api_key: str, file_url: str) -> dict:
+    """Fetch file_url bytes (NO api key on that GET), then POST as multipart ``file``.
+
+    Agent platforms hand tools a URL, not a local path, so the source fetch — and any
+    presigned result URL — are retrieved without the api-key header; only the EnConvert
+    POST carries X-API-Key. Inlined per component (bundle files are imported standalone).
+    """
+    if not api_key:
+        raise ValueError(
+            "EnConvert API key is missing. Add your private key (starts with sk_) "
+            "from https://www.enconvert.com/dashboard/api-keys."
+        )
+    if not file_url:
+        raise ValueError("Provide a file URL to convert.")
+    src = requests.get(file_url, timeout=TIMEOUT)  # no api-key header on the source fetch
+    src.raise_for_status()
+    filename = file_url.split("?", 1)[0].rsplit("/", 1)[-1] or "file"
+    resp = requests.post(
+        BASE_URL + endpoint,
+        files={"file": (filename, io.BytesIO(src.content))},
+        headers={"X-API-Key": api_key},
+        timeout=TIMEOUT,
+    )
+    if resp.status_code in (401, 403):
+        raise ValueError(
+            f"EnConvert rejected the API key (HTTP {resp.status_code}). Use a private "
+            "key (sk_...); public pk_ keys are not accepted."
+        )
+    resp.raise_for_status()
+    return resp.json()
+
+
+class EnConvertConvertToMarkdown(Component):
+    display_name = "Convert File to Markdown"
+    description = (
+        "Convert any file (from a URL) into clean markdown. The file bytes are fetched "
+        "from the URL, converted, and returned as a presigned download URL."
+    )
+    documentation = "https://www.enconvert.com/docs"
+    icon = "FileText"
+    name = "enconvert_convert_to_markdown"  # stable internal id — do not rename
+
+    inputs = [
+        MessageTextInput(
+            name="file_url",
+            display_name="File URL",
+            required=True,
+            info="URL of the file to convert. Fetched without your api key.",
+        ),
+        SecretStrInput(
+            name="api_key",
+            display_name="EnConvert API Key",
+            required=True,
+            info="Private key (sk_...) from https://www.enconvert.com/dashboard/api-keys.",
+        ),
+    ]
+
+    outputs = [Output(display_name="Data", name="data", method="build")]
+
+    def build(self) -> Data:
+        result = _convert("/v1/convert/anything-to-markdown", self.api_key, self.file_url)
+        self.status = result.get("presigned_url", "done")
+        return Data(data=result)

lfx_enconvert-0.0.1/src/lfx_enconvert/components/enconvert/convert_pdf.py

@@ -0,0 +1,78 @@
+"""Convert File to PDF — any file URL into a PDF."""
+
+from __future__ import annotations
+
+import io
+
+import requests
+from lfx.custom.custom_component.component import Component
+from lfx.io import MessageTextInput, Output, SecretStrInput
+from lfx.schema.data import Data
+
+BASE_URL = "https://api.enconvert.com"
+TIMEOUT = 120
+
+
+def _convert(endpoint: str, api_key: str, file_url: str) -> dict:
+    """Fetch file_url bytes (NO api key on that GET), then POST as multipart ``file``.
+
+    Agent platforms hand tools a URL, not a local path, so the source fetch — and any
+    presigned result URL — are retrieved without the api-key header; only the EnConvert
+    POST carries X-API-Key. Inlined per component (bundle files are imported standalone).
+    """
+    if not api_key:
+        raise ValueError(
+            "EnConvert API key is missing. Add your private key (starts with sk_) "
+            "from https://www.enconvert.com/dashboard/api-keys."
+        )
+    if not file_url:
+        raise ValueError("Provide a file URL to convert.")
+    src = requests.get(file_url, timeout=TIMEOUT)  # no api-key header on the source fetch
+    src.raise_for_status()
+    filename = file_url.split("?", 1)[0].rsplit("/", 1)[-1] or "file"
+    resp = requests.post(
+        BASE_URL + endpoint,
+        files={"file": (filename, io.BytesIO(src.content))},
+        headers={"X-API-Key": api_key},
+        timeout=TIMEOUT,
+    )
+    if resp.status_code in (401, 403):
+        raise ValueError(
+            f"EnConvert rejected the API key (HTTP {resp.status_code}). Use a private "
+            "key (sk_...); public pk_ keys are not accepted."
+        )
+    resp.raise_for_status()
+    return resp.json()
+
+
+class EnConvertConvertToPdf(Component):
+    display_name = "Convert File to PDF"
+    description = (
+        "Convert any file (from a URL) into a PDF. The file bytes are fetched from the "
+        "URL, converted, and returned as a presigned download URL."
+    )
+    documentation = "https://www.enconvert.com/docs"
+    icon = "File"
+    name = "enconvert_convert_to_pdf"  # stable internal id — do not rename
+
+    inputs = [
+        MessageTextInput(
+            name="file_url",
+            display_name="File URL",
+            required=True,
+            info="URL of the file to convert. Fetched without your api key.",
+        ),
+        SecretStrInput(
+            name="api_key",
+            display_name="EnConvert API Key",
+            required=True,
+            info="Private key (sk_...) from https://www.enconvert.com/dashboard/api-keys.",
+        ),
+    ]
+
+    outputs = [Output(display_name="Data", name="data", method="build")]
+
+    def build(self) -> Data:
+        result = _convert("/v1/convert/anything-to-pdf", self.api_key, self.file_url)
+        self.status = result.get("presigned_url", "done")
+        return Data(data=result)

lfx_enconvert-0.0.1/src/lfx_enconvert/components/enconvert/discover_urls.py

@@ -0,0 +1,107 @@
+"""Discover URLs — map or crawl a site into a list of URLs."""
+
+from __future__ import annotations
+
+import requests
+from lfx.custom.custom_component.component import Component
+from lfx.io import BoolInput, DropdownInput, IntInput, MessageTextInput, Output, SecretStrInput
+from lfx.schema.data import Data
+
+BASE_URL = "https://api.enconvert.com"
+TIMEOUT = 120
+
+
+def _post(endpoint: str, api_key: str, payload: dict) -> dict:
+    """POST JSON to an EnConvert endpoint with the private-key header, return parsed JSON.
+
+    Inlined per component: the Langflow loader imports bundle files standalone, so
+    relative imports between them are unsupported.
+    """
+    if not api_key:
+        raise ValueError(
+            "EnConvert API key is missing. Add your private key (starts with sk_) "
+            "from https://www.enconvert.com/dashboard/api-keys."
+        )
+    resp = requests.post(
+        BASE_URL + endpoint, json=payload, headers={"X-API-Key": api_key}, timeout=TIMEOUT
+    )
+    if resp.status_code in (401, 403):
+        raise ValueError(
+            f"EnConvert rejected the API key (HTTP {resp.status_code}). Use a private "
+            "key (sk_...); public pk_ keys are not accepted."
+        )
+    resp.raise_for_status()
+    return resp.json()
+
+
+def _csv(raw: str) -> list[str]:
+    return [s.strip() for s in raw.split(",") if s.strip()] if raw else []
+
+
+class EnConvertDiscoverUrls(Component):
+    display_name = "Discover URLs"
+    description = "Map a site via sitemap, crawl, or hybrid, and return the discovered URLs."
+    documentation = "https://www.enconvert.com/docs"
+    icon = "Compass"
+    name = "enconvert_discover_urls"  # stable internal id — do not rename
+
+    inputs = [
+        MessageTextInput(
+            name="url",
+            display_name="URL",
+            required=True,
+            info="The site to discover URLs from.",
+        ),
+        DropdownInput(
+            name="mode",
+            display_name="Mode",
+            options=["hybrid", "sitemap", "crawl"],
+            value="hybrid",
+        ),
+        IntInput(name="max_urls", display_name="Max URLs", value=100),
+        IntInput(name="max_depth", display_name="Max depth", value=2),
+        MessageTextInput(
+            name="include_patterns",
+            display_name="Include patterns",
+            info="Optional comma-separated glob patterns to keep.",
+            advanced=True,
+        ),
+        MessageTextInput(
+            name="exclude_patterns",
+            display_name="Exclude patterns",
+            info="Optional comma-separated glob patterns to drop.",
+            advanced=True,
+        ),
+        BoolInput(
+            name="same_domain_only",
+            display_name="Same domain only",
+            value=True,
+            advanced=True,
+        ),
+        SecretStrInput(
+            name="api_key",
+            display_name="EnConvert API Key",
+            required=True,
+            info="Private key (sk_...) from https://www.enconvert.com/dashboard/api-keys.",
+        ),
+    ]
+
+    outputs = [Output(display_name="Data", name="data", method="build")]
+
+    def build(self) -> Data:
+        payload: dict = {
+            "url": self.url,
+            "mode": self.mode,
+            "max_urls": self.max_urls,
+            "max_depth": self.max_depth,
+            "same_domain_only": self.same_domain_only,
+        }
+        include = _csv(self.include_patterns)
+        exclude = _csv(self.exclude_patterns)
+        if include:
+            payload["include_patterns"] = include
+        if exclude:
+            payload["exclude_patterns"] = exclude
+        result = _post("/v2/discover", self.api_key, payload)
+        self.status = f"{result.get('total', len(result.get('urls', [])))} URLs"
+        return Data(data=result)

lfx_enconvert-0.0.1/src/lfx_enconvert/components/enconvert/extract_structured.py

@@ -0,0 +1,110 @@
+"""Extract Structured (distill) — pull typed data from pages against a schema."""
+
+from __future__ import annotations
+
+import json
+
+import requests
+from lfx.custom.custom_component.component import Component
+from lfx.io import IntInput, MessageTextInput, MultilineInput, Output, SecretStrInput
+from lfx.schema.data import Data
+
+BASE_URL = "https://api.enconvert.com"
+TIMEOUT = 120
+
+
+def _post(endpoint: str, api_key: str, payload: dict) -> dict:
+    """POST JSON to an EnConvert endpoint with the private-key header, return parsed JSON.
+
+    Inlined per component: the Langflow loader imports bundle files standalone, so
+    relative imports between them are unsupported.
+    """
+    if not api_key:
+        raise ValueError(
+            "EnConvert API key is missing. Add your private key (starts with sk_) "
+            "from https://www.enconvert.com/dashboard/api-keys."
+        )
+    resp = requests.post(
+        BASE_URL + endpoint, json=payload, headers={"X-API-Key": api_key}, timeout=TIMEOUT
+    )
+    if resp.status_code in (401, 403):
+        raise ValueError(
+            f"EnConvert rejected the API key (HTTP {resp.status_code}). Use a private "
+            "key (sk_...); public pk_ keys are not accepted."
+        )
+    resp.raise_for_status()
+    return resp.json()
+
+
+def _csv(raw: str) -> list[str]:
+    return [s.strip() for s in raw.split(",") if s.strip()] if raw else []
+
+
+class EnConvertExtractStructured(Component):
+    display_name = "Extract Structured"
+    description = (
+        "Extract structured data from one or more pages against a JSON schema "
+        "(or a flat field:description map). Give either URLs or a site to discover from."
+    )
+    documentation = "https://www.enconvert.com/docs"
+    icon = "Braces"
+    name = "enconvert_extract_structured"  # stable internal id — do not rename
+
+    inputs = [
+        MultilineInput(
+            name="extraction_schema",
+            display_name="Schema",
+            required=True,
+            info=(
+                'A JSON schema {"type":"object","properties":{...}} OR a flat '
+                '{"field":"description"} map.'
+            ),
+        ),
+        MessageTextInput(
+            name="urls",
+            display_name="URLs",
+            info="Comma-separated page URLs (max 50). Provide this OR 'Discover from URL'.",
+        ),
+        MessageTextInput(
+            name="discover_from_url",
+            display_name="Discover from URL",
+            info="A site to discover pages from, instead of listing URLs.",
+        ),
+        IntInput(
+            name="discover_max_pages",
+            display_name="Discover max pages",
+            value=25,
+            advanced=True,
+            info="Only used with 'Discover from URL'.",
+        ),
+        SecretStrInput(
+            name="api_key",
+            display_name="EnConvert API Key",
+            required=True,
+            info="Private key (sk_...) from https://www.enconvert.com/dashboard/api-keys.",
+        ),
+    ]
+
+    outputs = [Output(display_name="Data", name="data", method="build")]
+
+    def build(self) -> Data:
+        urls = _csv(self.urls)
+        discover = (self.discover_from_url or "").strip()
+        if bool(urls) == bool(discover):
+            raise ValueError(
+                "Provide exactly one of 'URLs' or 'Discover from URL' (not both, not neither)."
+            )
+        raw_schema = (self.extraction_schema or "").strip()
+        if not raw_schema:
+            raise ValueError("Provide a JSON schema or a flat {field: description} map.")
+        payload: dict = {"schema": json.loads(raw_schema)}
+        if urls:
+            payload["urls"] = urls
+        else:
+            discover_from: dict = {"url": discover}
+            if self.discover_max_pages:
+                discover_from["max_pages"] = self.discover_max_pages
+            payload["discover_from"] = discover_from
+        result = _post("/v2/distill", self.api_key, payload)
+        self.status = f"tier={result.get('extraction_tier')}"
+        return Data(data=result)

lfx_enconvert-0.0.1/src/lfx_enconvert/components/enconvert/perceive.py

@@ -0,0 +1,93 @@
+"""Perceive URL — render a web page into agent-ready outputs with a quality score."""
+
+from __future__ import annotations
+
+import requests
+from lfx.custom.custom_component.component import Component
+from lfx.io import BoolInput, MessageTextInput, Output, SecretStrInput
+from lfx.schema.data import Data
+
+BASE_URL = "https://api.enconvert.com"
+TIMEOUT = 120
+
+
+def _post(endpoint: str, api_key: str, payload: dict) -> dict:
+    """POST JSON to an EnConvert endpoint with the private-key header, return parsed JSON.
+
+    Bundle modules are imported standalone by the Langflow loader, so this helper is
+    inlined per component (relative imports between bundle files are unsupported).
+    """
+    if not api_key:
+        raise ValueError(
+            "EnConvert API key is missing. Add your private key (starts with sk_) "
+            "from https://www.enconvert.com/dashboard/api-keys."
+        )
+    resp = requests.post(
+        BASE_URL + endpoint, json=payload, headers={"X-API-Key": api_key}, timeout=TIMEOUT
+    )
+    if resp.status_code in (401, 403):
+        raise ValueError(
+            f"EnConvert rejected the API key (HTTP {resp.status_code}). Use a private "
+            "key (sk_...); public pk_ keys are not accepted."
+        )
+    resp.raise_for_status()
+    return resp.json()
+
+
+def _csv(raw: str) -> list[str]:
+    return [s.strip() for s in raw.split(",") if s.strip()] if raw else []
+
+
+class EnConvertPerceive(Component):
+    display_name = "Perceive URL"
+    description = (
+        "Render a web page into agent-ready outputs (markdown, structured data, "
+        "screenshot, PDF, links...) with a render_quality honesty score on every read."
+    )
+    documentation = "https://www.enconvert.com/docs"
+    icon = "Eye"
+    name = "enconvert_perceive"  # stable internal id — do not rename
+
+    inputs = [
+        MessageTextInput(
+            name="url",
+            display_name="URL",
+            required=True,
+            info="The page to perceive.",
+        ),
+        MessageTextInput(
+            name="formats",
+            display_name="Outputs",
+            value="markdown,structured",
+            info=(
+                "Comma-separated. Allowed: markdown, html_cleaned, html_raw, screenshot, "
+                "screenshot_full_page, pdf, links, images, structured. markdown and "
+                "structured come back inline; screenshot/pdf/html return 15-minute signed URLs."
+            ),
+        ),
+        BoolInput(
+            name="only_main_content",
+            display_name="Only main content",
+            value=False,
+            info="Strip navigation and boilerplate, keep the main content.",
+        ),
+        SecretStrInput(
+            name="api_key",
+            display_name="EnConvert API Key",
+            required=True,
+            info="Private key (sk_...) from https://www.enconvert.com/dashboard/api-keys.",
+        ),
+    ]
+
+    outputs = [Output(display_name="Data", name="data", method="build")]
+
+    def build(self) -> Data:
+        payload: dict = {"url": self.url}
+        formats = _csv(self.formats)
+        if formats:
+            payload["outputs"] = formats
+        if self.only_main_content:
+            payload["only_main_content"] = True
+        result = _post("/v2/perceive", self.api_key, payload)
+        self.status = f"render_quality={result.get('render_quality')}"
+        return Data(data=result)

lfx_enconvert-0.0.1/src/lfx_enconvert/components/enconvert/web_search.py

@@ -0,0 +1,104 @@
+"""Web Search — search the web (or news/images/scholar/patents/maps) via EnConvert."""
+
+from __future__ import annotations
+
+import requests
+from lfx.custom.custom_component.component import Component
+from lfx.io import DropdownInput, IntInput, MessageTextInput, Output, SecretStrInput
+from lfx.schema.data import Data
+
+BASE_URL = "https://api.enconvert.com"
+TIMEOUT = 120
+
+
+def _post(endpoint: str, api_key: str, payload: dict) -> dict:
+    """POST JSON to an EnConvert endpoint with the private-key header, return parsed JSON.
+
+    Inlined per component: the Langflow loader imports bundle files standalone, so
+    relative imports between them are unsupported.
+    """
+    if not api_key:
+        raise ValueError(
+            "EnConvert API key is missing. Add your private key (starts with sk_) "
+            "from https://www.enconvert.com/dashboard/api-keys."
+        )
+    resp = requests.post(
+        BASE_URL + endpoint, json=payload, headers={"X-API-Key": api_key}, timeout=TIMEOUT
+    )
+    if resp.status_code in (401, 403):
+        raise ValueError(
+            f"EnConvert rejected the API key (HTTP {resp.status_code}). Use a private "
+            "key (sk_...); public pk_ keys are not accepted."
+        )
+    resp.raise_for_status()
+    return resp.json()
+
+
+class EnConvertWebSearch(Component):
+    display_name = "Web Search"
+    description = "Search the web and get back structured results (title, url, snippet, position)."
+    documentation = "https://www.enconvert.com/docs"
+    icon = "Search"
+    name = "enconvert_web_search"  # stable internal id — do not rename
+
+    inputs = [
+        MessageTextInput(
+            name="query",
+            display_name="Query",
+            required=True,
+            info="The search query.",
+        ),
+        DropdownInput(
+            name="category",
+            display_name="Category",
+            options=["web", "news", "images", "scholar", "patents", "maps"],
+            value="web",
+        ),
+        IntInput(
+            name="num_results",
+            display_name="Results",
+            value=10,
+            info="Number of results to return.",
+        ),
+        MessageTextInput(
+            name="country",
+            display_name="Country",
+            info="Optional ISO country code (e.g. us).",
+            advanced=True,
+        ),
+        MessageTextInput(
+            name="locale",
+            display_name="Locale",
+            info="Optional locale (e.g. en).",
+            advanced=True,
+        ),
+        DropdownInput(
+            name="time_filter",
+            display_name="Time filter",
+            options=["", "hour", "day", "week", "month", "year"],
+            value="",
+            advanced=True,
+        ),
+        SecretStrInput(
+            name="api_key",
+            display_name="EnConvert API Key",
+            required=True,
+            info="Private key (sk_...) from https://www.enconvert.com/dashboard/api-keys.",
+        ),
+    ]
+
+    outputs = [Output(display_name="Data", name="data", method="build")]
+
+    def build(self) -> Data:
+        payload: dict = {"query": self.query, "category": self.category}
+        if self.num_results:
+            payload["num_results"] = self.num_results
+        if self.country:
+            payload["country"] = self.country
+        if self.locale:
+            payload["locale"] = self.locale
+        if self.time_filter:
+            payload["time_filter"] = self.time_filter
+        result = _post("/v2/search", self.api_key, payload)
+        self.status = f"{len(result.get('results', []))} results"
+        return Data(data=result)

lfx_enconvert-0.0.1/src/lfx_enconvert/extension.json

@@ -0,0 +1,16 @@
+{
+  "$schema": "https://schemas.langflow.org/extension/v1.json",
+  "id": "lfx-enconvert",
+  "version": "0.0.1",
+  "name": "EnConvert",
+  "description": "EnConvert bundle for Langflow: perceive URLs, search the web, discover URLs, extract structured data, and convert files — every read scored with render_quality.",
+  "lfx": {
+    "compat": ["1"]
+  },
+  "bundles": [
+    {
+      "name": "enconvert",
+      "path": "components/enconvert"
+    }
+  ]
+}