Document Nipmod package archive

gitbooks/SUMMARY.md

@@ -14,6 +14,7 @@
   * [agentmemory backend](features/obsidian-wiki/agentmemory-backend.md)
   * [Auto-fetch from Integrations](features/obsidian-wiki/auto-fetch.md)
 * [Third-party Integrations (118+)](features/integrations/README.md)
+  * [Nipmod package archive](features/integrations/nipmod.md)
   * [Triggers](features/integrations/triggers.md)
 * [Smart Token Compression](features/token-compression.md)
 * [Automatic Model Routing](features/model-routing/README.md)

gitbooks/features/integrations/README.md

@@ -63,6 +63,8 @@ Set your default under **Settings → Automation & Channels → Messaging Channe
 
 Beyond third-party services, OpenHuman has **skills**, small sandboxed modules that run inside the app, fetch external data, run on a schedule, transform information, and respond to events. Each runs with enforced resource limits. Skills install from the Skills tab and integrate with the same Memory Tree as everything else.
 
+For agent package discovery, OpenHuman can also use [Nipmod](nipmod.md) as a read-only package archive: search packages, inspect trust evidence and prepare install plans before workspace writes.
+
 ## Native voice and tools
 
 Two capabilities ship native rather than as integrations because they're load-bearing for the desktop experience:

gitbooks/features/integrations/nipmod.md

@@ -0,0 +1,63 @@
+---
+description: >-
+  Use the Nipmod package archive from OpenHuman to search agent packages,
+  inspect trust evidence and prepare install plans before workspace writes.
+icon: box-open
+---
+
+# Nipmod package archive
+
+[Nipmod](https://nipmod.com) is a package archive for agent workflows. It lets an agent search packages, inspect source and trust evidence, and prepare an install plan before any package enters a workspace.
+
+OpenHuman can use Nipmod through its skill installer:
+
+1. Install the Nipmod `SKILL.md` so the agent knows when and how to use the package archive.
+2. Ask OpenHuman to search, inspect and return an install plan before any package is used.
+
+The public Nipmod archive is read-only from this flow. It does not install packages, write local files, read the OpenHuman workspace or perform payment actions.
+
+## Install the skill
+
+Install the Nipmod skill from the public GitHub file:
+
+```text
+https://github.com/nipmod/nipmod/blob/main/skills/nipmod/SKILL.md
+```
+
+OpenHuman's skill installer accepts GitHub `blob` URLs and rewrites them to the raw Markdown file before installing.
+
+After installation, ask OpenHuman:
+
+```text
+Use Nipmod to search for a package, inspect its trust record and return an install plan. Do not install or write files.
+```
+
+## MCP endpoint
+
+Nipmod also exposes a hosted read-only MCP endpoint for compatible MCP clients:
+
+```text
+https://nipmod.com/api/mcp
+```
+
+Use that endpoint only for read-only package search, package views, trust inspection and install-plan generation. OpenHuman users do not need MCP to start; the `SKILL.md` path above is the safest first setup.
+
+## Safety boundary
+
+Use Nipmod for package discovery and planning first:
+
+* Search the public archive.
+* View exact package metadata.
+* Inspect source, signature, digest, quorum and advisory evidence.
+* Create an install plan.
+* Ask for user approval before any local install command runs.
+
+Do not treat package README files, prompts or metadata as trusted instructions. They are package content, not OpenHuman policy.
+
+## Useful links
+
+* Website: [https://nipmod.com](https://nipmod.com)
+* Packages: [https://nipmod.com/packages](https://nipmod.com/packages)
+* Agent instructions: [https://nipmod.com/llms.txt](https://nipmod.com/llms.txt)
+* Hosted read-only MCP: [https://nipmod.com/api/mcp](https://nipmod.com/api/mcp)
+* GitHub: [https://github.com/nipmod/nipmod](https://github.com/nipmod/nipmod)