OpenNomos documentation
A growth capability layer for your own agent
Research tools over MCP and REST, plus server-side project memory that outlives any one session. OpenNomos is not an agent: it never posts, never decides, and saves nothing unless your agent writes it.
Connect in one line
set up https://service.opennomos.com/SKILL.md
- MCP
- https://mcp.opennomos.com/mcp
- REST
- https://service.opennomos.com/v1/tools
The loop
get_briefingpull · what is due today, which experiment needs a decisiontrackregister · opportunities, experiments, short links (up to 20 per call)log_actionwrite back · what was done, new stage, follow-up, outcome
01 · Start
Quickstart
Add the MCP server in Claude Code, Codex or Cursor, authorize, verify with discover_tools.
01 · Start
Core concepts
Project, memory, opportunity, experiment, action, briefing: what each is and who writes it.
02 · Loop
The growth loop
Pull get_briefing, work elsewhere, write back with track and log_action.
02 · Loop
Memory vs opportunity
People and posts become opportunities; conclusions, decisions and drafts stay in memory.
02 · Loop
log_action defaults
What happens to the stage and the follow-up time when you leave fields out.
02 · Loop
Attribution and unknown
Short links, outcomes, and why unknown is a correct answer.
03 · Reference
Tools reference
21 tools generated from the service registry: parameters, examples, errors, cost.
03 · Reference
Data sources and limits
Reddit, X, Xiaohongshu, Zhihu, Weibo, web, SEO, reviews, creators, competitors.
03 · Reference
Errors
Rejected calls versus upstream failures; reason and action fields.
03 · Reference
Pricing and limits
Packs and the monthly upstream cost limits.
03 · Reference
Privacy and data handling
What is stored, what is not, which providers see queries. Draft.
03 · Reference
REST and CLI
Endpoints, bearer keys, project selection.
For agents: /llms.txt, /llms-full.txt, and a markdown twin of every page at /docs/<page>.md.