wiki-history-ingest Setup Guide¶
Skill: wiki-history-ingest · Installs: 2,744 · Source: ar9av/obsidian-wiki
Unified entrypoint for ingesting agent conversation history into an Obsidian wiki vault. Routes /wiki-history-ingest hermes and /wiki-history-ingest openclaw to their respective specialized history ingestion skills.
Prerequisites¶
- Obsidian vault set up with the obsidian-wiki plugin structure
- Python 3.10+ for the ingestion scripts
- Agent history files accessible at standard paths:
- Hermes:
~/.hermes/(memories, sessions) - OpenClaw:
~/.openclaw/(MEMORY.md, session JSONL) - Claude:
~/.claude/(memory/session JSONL) - Codex:
~/.codex/(rollout/session index) - Copilot:
~/.copilot/orsession-store.db
Installation¶
npx skills add ar9av/obsidian-wiki --skill wiki-history-ingest
The skill installs to your agent's skills directory. It works as a thin router - it dispatches to destination skills (hermes-history-ingest, openclaw-history-ingest, etc.) which must also be installed.
Install destination skills¶
# For Hermes agent history
npx skills add ar9av/obsidian-wiki --skill hermes-history-ingest
# For OpenClaw agent history
npx skills add ar9av/obsidian-wiki --skill openclaw-history-ingest
# For Claude, Codex, Copilot, Pi (as needed)
npx skills add ar9av/obsidian-wiki --skill claude-history-ingest
npx skills add ar9av/obsidian-wiki --skill codex-history-ingest
npx skills add ar9av/obsidian-wiki --skill copilot-history-ingest
npx skills add ar9av/obsidian-wiki --skill pi-history-ingest
Capabilities¶
| Subcommand | Routes To | Source Path Auto-Detection |
|---|---|---|
hermes |
hermes-history-ingest |
~/.hermes memories/session artifacts |
openclaw |
openclaw-history-ingest |
~/.openclaw MEMORY.md / session JSONL |
claude |
claude-history-ingest |
~/.claude memory/session JSONL |
codex |
codex-history-ingest |
~/.codex rollout/session index |
copilot |
copilot-history-ingest |
~/.copilot session-store.db |
pi |
pi-history-ingest |
~/.pi/agent/sessions |
auto |
Infers from context | Uses source path detection rules |
Usage¶
Basic: Ingest Hermes history¶
/wiki-history-ingest hermes
This routes to hermes-history-ingest which processes Hermes memories and session data into structured Obsidian notes.
Auto-detect from path¶
/wiki-history-ingest auto ~/.hermes/profiles/<your-profile>/
The router detects ~/.hermes in the path and dispatches to hermes-history-ingest.
Ingest all agent histories¶
/wiki-history-ingest claude
/wiki-history-ingest codex
/wiki-history-ingest hermes
/wiki-history-ingest openclaw
CorpusIQ Use Cases¶
Agent session archival: After each Hermes session, ingest the conversation history into the CorpusIQ knowledge wiki for long-term search and retrieval.
Cross-agent context synthesis: Ingest history from Hermes, Claude Code, and Codex into a unified vault - then use wiki-synthesize to find patterns across agent sessions.
Compliance and audit: Maintain a searchable archive of all agent interactions for governance and audit requirements.
Troubleshooting¶
| Issue | Fix |
|---|---|
Skill not found: hermes-history-ingest |
Install the destination skill first: npx skills add ar9av/obsidian-wiki --skill hermes-history-ingest |
No such file: ~/.hermes |
Verify Hermes is installed and has generated session data at the standard path |
| Router dispatches to wrong skill | Use explicit subcommand instead of auto - e.g., /wiki-history-ingest hermes |
| Empty vault after ingest | Check that the destination skill's Python dependencies are installed (pip install -r requirements.txt in the obsidian-wiki repo) |
Related Skills¶
- hermes-history-ingest - Hermes-specific history ingestion
- openclaw-history-ingest - OpenClaw-specific history ingestion
- wiki-synthesize - Cross-session pattern synthesis
- wiki-query - Natural language vault queries
← Skills Marketplace | Skills Catalog → Curated by CorpusIQ - one MCP endpoint, all your business tools.
Integration with CorpusIQ¶
CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.
This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.