Agents
The llm-wiki skill uses the standard agentskills.io format, so it runs in every agent below. The wiki operations are available through the installed skill: init/runtime setup, hybrid or lexical search, structural lint, stats, graph lint, graph extraction, and graph query. Dependency-bearing scripts use pinned PEP 723 environments through uv run --script.
Support matrix
| Agent | --agent value | Invoke via | Scripts run |
|---|---|---|---|
| Claude Code | claude-code | /wiki:* slash commands (bundled) or natural language | ✅ |
| Codex (OpenAI) | codex | /skills or $llm-wiki / natural language | ✅ |
| Cursor | cursor | /llm-wiki or natural language | ✅ |
| Gemini CLI | gemini-cli | /skills management commands / natural language | ✅ |
| OpenCode | opencode | natural language (agent invokes the native skill tool) | ✅ |
| OpenClaw | openclaw | auto-exposed as a user command | ✅ |
| Pi Agent | pi | /skill:llm-wiki or natural language | ✅ |
| OMP ("Oh My Pi") | manual (see below) | natural language (skills auto-surface via skill://) | ✅ |
| Hermes Agent | manual (see below) | natural language | ✅ |
OpenCode also reads .claude/skills/ and ~/.claude/skills/, so a Claude Code install works there without a second install.
Install snippets
Install uv first. Every agent invokes the same pinned local runtime; no agent-specific Python environment is required.
Claude Code gets the full plugin (skill + slash commands + manifest):
/plugin marketplace add praneybehl/llm-wiki-plugin
/plugin install llm-wiki@llm-wikiCodex, Cursor, Gemini CLI, OpenCode, OpenClaw, Pi install the skill through the skills CLI — substitute the matrix's --agent value:
npx skills add praneybehl/llm-wiki-plugin -a codex -g
npx skills add praneybehl/llm-wiki-plugin -a cursor -g
npx skills add praneybehl/llm-wiki-plugin -a gemini-cli -g
npx skills add praneybehl/llm-wiki-plugin -a opencode -g
npx skills add praneybehl/llm-wiki-plugin -a openclaw -g
npx skills add praneybehl/llm-wiki-plugin -a pi -gDrop -g to install into the current project only.
OMP ("Oh My Pi") and Hermes Agent aren't yet in the npx skills registry — clone the repo and symlink the skill into the agent's skills directory. OMP reads managed/user skills from ~/.omp/agent/skills/ and surfaces them via skill://:
git clone https://github.com/praneybehl/llm-wiki-plugin.git
mkdir -p ~/.omp/agent/skills ~/.hermes/skills
ln -s "$(pwd)/llm-wiki-plugin/skills/llm-wiki" ~/.omp/agent/skills/llm-wiki
ln -s "$(pwd)/llm-wiki-plugin/skills/llm-wiki" ~/.hermes/skills/llm-wikiAgent-memory file
After a wiki is bootstrapped, add a short stanza so the agent remembers its location. Put a stable user-level path such as ~/wiki/ in the agent's global instructions for one personal wiki across projects, or a relative path in the project's memory file for a project-only wiki. The filename depends on the agent:
| Agent | Memory file |
|---|---|
| Claude Code | CLAUDE.md |
| Codex (OpenAI) | AGENTS.md |
| Cursor | AGENTS.md (or .cursor/rules/*.mdc) |
| OpenCode | AGENTS.md (also reads CLAUDE.md) |
| Gemini CLI | GEMINI.md |
| Pi Agent | AGENTS.md |
| OpenClaw | AGENTS.md |
| OMP (Oh My Pi) | AGENTS.md |
Running multiple agents? Prefer AGENTS.md as the canonical file where supported and symlink CLAUDE.md to it. The skill never writes a memory file without your approval. Global and project stanza variants live in skills/llm-wiki/references/agent-memory-integration.md.
INFO
Slash commands are Claude Code–only. Everywhere else, invoke the skill by natural language: "add this paper to the wiki", "what does the wiki say about X", "lint the wiki". The wiki itself is agent-agnostic — ingest with one agent, query with another.
Experience and procedure improvements
All listed skill-compatible agents can run /wiki:learn and /wiki:evolve workflows through natural language using the installed references/evolution-workflow.md and scripts/wiki_evolve.py. Only Claude Code has the literal slash commands. Capture and lifecycle tooling use Python stdlib. The shared evaluation adapter includes all nine hosts. Their installed versions, authentication, permitted tools and reported accounting differ; see the evolution execution requirements and live-study report. No new agent integration is required to capture lessons. See Learning and skill evolution.