Connapse
llms.txt22 starsChanged 5 days agoMITView on GitHub
Connapse is a self-hosted knowledge platform. It indexes uploaded files and external sources (Amazon S3, Azure Blob Storage, GitHub, Confluence, Jira, SFTP and custom source servers) into containers, and serves hybrid vector and keyword search over a Blazor web UI, a REST API at /api/v1 and an MCP server at /mcp. Built on .NET 10, PostgreSQL with pgvector, and MinIO. Every page below is Markdown. API keys go in the X-Api-Key header; errors are RFC 9457 problem details whose type…
In this repository6 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- create-ticketsSkill
- discover-workSkill
- next-taskSkill
- parallel-workSkill
- release-gatekeeperSkill
The file
We split this llms.txt into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 5
Every page below is Markdown.
≈ “Common” is an estimate: at least 1 in 5 of the 3,592 llms.txt files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
# Connapse > Connapse is a self-hosted knowledge platform. It indexes uploaded files and external sources (Amazon S3, Azure Blob Storage, GitHub, Confluence, Jira, SFTP and custom source servers) into containers, and serves hybrid vector and keyword search over a Blazor web UI, a REST API at /api/v1 and an MCP server at /mcp. Built on .NET 10, PostgreSQL with pgvector, and MinIO. Every page below is Markdown. API keys go in the `X-Api-Key` header; errors are RFC 9457 problem details whose `type` links into problems.md. ## Get started - [Quickstart](get-started/quickstart.md): run Connapse with Docker Compose, upload a document and search it - [Tutorial: connect Claude to Connapse](get-started/connect-an-agent.md): create an agent key, connect over MCP, build a container and save a memory ## Self-host - [Install Connapse](self-host/install.md): requirements and Compose profiles (Ollama, reranker, GPU) - [Run Connapse in production](self-host/production.md): TLS, secrets, reverse proxy and hardening - [Back up and restore Connapse](self-host/backup-and-restore.md): PostgreSQL, MinIO and the data directory - [Upgrade Connapse](self-host/upgrade.md): upgrade, migrations, rollback and the v0.3.x to v0.4.0 steps - [Troubleshoot a self-hosted server](self-host/troubleshooting.md): symptoms and fixes ## Sources - [Containers, connections and sources](sources/index.md): how they fit together and what each provider indexes - [Use uploads and MinIO](sources/uploads-and-minio.md): the default managed storage - [Index a folder over SFTP](sources/sftp.md): folders on your computer, a NAS or another server - [AWS setup](sources/aws-setup.md): Amazon S3 with per-user S3 Access Grants - [Cover buckets outside the Identity Center region](sources/aws-multi-region.md): multi-region AWS - [Azure setup](sources/azure-setup.md): Azure Blob Storage with per-user Entra permissions - [GitHub setup](sources/github-setup.md): repository docs, issues and pull requests - [Atlassian setup](sources/atlassian-setup.md): Confluence spaces and Jira projects - [Custom source setup](sources/custom-source-setup.md): connect your own system through a custom source server ## Guides - [Manage users and invites](guides/users-and-invites.md): roles, invites and removing users - [Create tokens and agent keys](guides/tokens-and-agents.md): personal access tokens, scopes and agent keys - [Connect an MCP client](guides/mcp-clients.md): Claude Code, Claude Desktop, VS Code, Cursor and approvals - [Automate Connapse with the API](guides/automate-with-the-api.md): providers, connections and sources over REST, with ETag saves - [Tune search](guides/tune-search.md): providers, reranker, fusion and chunking - [Change the embedding model](guides/change-embedding-model.md): switch models and re-index - [Search code with an agent](guides/search-code.md): index C#, TypeScript and Python and use the code_* MCP tools ## Concepts - [How Connapse works](concepts/how-connapse-works.md): the parts and how a document becomes a search result - [About search](concepts/search.md): hybrid search, fusion, reranking and AutoCut - [Following links](concepts/search.md#following-links): link search across GitHub, Jira, Confluence and code (EnableLinkSearch, off by default) - [About ingestion](concepts/ingestion.md): parsing, chunking and reindexing - [About permissions](concepts/permissions.md): per-object trimming, cloud identity and failing closed - [About agent memory](concepts/memory.md): dated facts with sources, who sees them, corrections and forgetting ## Reference - [REST API](reference/rest-api.md): every endpoint with its scope, plus auth, errors, paging and rate limits - [OpenAPI document](api/openapi.json): the machine-readable API - [MCP tools](reference/mcp-tools.md): the 21 named tools (including code and agent memory tools) and 3 generic API tools - [Configuration](reference/configuration.md): every setting, environment variable and default - [Roles and scopes](reference/roles-and-scopes.md): which role and token scope each operation needs - [File types](reference/file-types.md): supported formats, limits and failure codes - [Error types](problems.md): every problem `type` the API returns - [Custom source protocol](protocols/custom-source-v1.md): the contract a custom source server implements - [Command-line tool](reference/cli.md): the connapse CLI ## Optional - [Set up a development environment](contributing/development.md): build, test and run migrations - [Architecture internals](contributing/architecture.md): projects, patterns and data flow
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

