Metadata-Version: 2.4
Name: p53-courier
Version: 0.1.0
Summary: CLI-based AI assisted web search with configurable breadth and depth
Author: Point 53 LLC
License-Expression: MPL-2.0
License-File: LICENSE
License-File: NOTICE
License-File: THIRD_PARTY_LICENSES.md
Keywords: ai,local-llm,ollama,point53,research,search,web
Classifier: Development Status :: 3 - Alpha
Classifier: Environment :: Console
Classifier: Intended Audience :: Developers
Classifier: Intended Audience :: End Users/Desktop
Classifier: License :: OSI Approved :: Mozilla Public License 2.0 (MPL 2.0)
Classifier: Operating System :: MacOS
Classifier: Operating System :: POSIX :: Linux
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.10
Classifier: Topic :: Internet :: WWW/HTTP
Classifier: Topic :: Text Processing :: Markup :: Markdown
Requires-Python: >=3.10
Requires-Dist: anthropic>=0.75.0
Requires-Dist: click>=8.3.1
Requires-Dist: dateparser>=1.2.2
Requires-Dist: feedparser>=6.0.12
Requires-Dist: httpx>=0.27.0
Requires-Dist: mcp>=1.0.0
Requires-Dist: ollama>=0.6.1
Requires-Dist: p53-collector
Requires-Dist: platformdirs>=4.0
Requires-Dist: protobuf>=6.33.2
Requires-Dist: pydantic>=2.0
Requires-Dist: selenium>=4.39.0
Requires-Dist: tomli>=2.0; python_version < '3.11'
Requires-Dist: webdriver-manager>=4.0.2
Description-Content-Type: text/markdown

# Point 53 Courier

**Courier** is a CLI-based AI-assisted web search tool — part of the Point 53 suite. It starts with a web search and, using any Ollama model(s) you configure, contributes analysis and summarization on top of the results.

## It All Starts With a Web Search

Then, using any Ollama model(s), you can configure which to contribute analysis and summarization to.

## XSS? I hardly know her.

Your LLM extracts XSS from the search query, to be analyzed by code. It goes through the first m results, then...

it goes n layers deep through the resulting pages with links, again with max m. n=2 is thought to be reasonable, 5 is risky, and 10 is probably dangerous. Use this technology at your own risk.

Well, we haven't tested it yet. We think that will be the case though. Enjoy!

## Using -p Flag for Passthrough

Run it through another layer of an LLM that uses up more context for extra analysis.

By default, it will summarize the outer-most edges of the search radius...
With Passthrough, it will go over all the RSS and create another RSS for additional information and go through that additional information as well.

## Straightforward breakdown.

"Your search Query"
|
|
|
Search results (include first m link data in context for sub-analysis)
- Completed by the rss-model name
- Automatically passes the m count extracted links most relevant to search query as an RSS.
- Selenium goes through each of these and we include the title, Link, site links, and site text in the context
(The above is ** ALWAYS ** output for context of original search compared to n-depth)
- NOTE: A summary should be provided to both humans in the final output and the LLMs here too; nobody wants to read sporadic and wacky site content.

Link 1  Link 2
|       |
|       |
|       |
|       |
If links are present, go to them and repeat JUST XSS creation n-1 times
    - Summarize on the last outer edge of the run, similar to results of the original search
    - Model choice will be important here
        - Need to figure out if separating by -p like informer program or merging with this section is the best
