# Ingest Document

> Base domain for extracted beliefs (optional) Use when this capability is needed.

- Skill: `tomevault-io/ingest-document` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add tomevault-io/ingest-document`
- Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/ingest-document/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: tomevault-io (https://skillmd.com/u/tomevault-io)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/tomevault-io/ingest-document

---


# Ingest Document

Process a document and extract beliefs from it into the knowledge substrate.

## Instructions

1. Read the document at the given path using the Read tool
2. Analyze the content to identify:
   - Factual claims and assertions
   - Decisions that were made
   - Preferences expressed
   - Technical specifications
   - Key entities (people, tools, projects)
3. For each extracted belief:
   - Assess confidence based on how clearly stated it is
   - Determine appropriate domain path
   - Identify related entities
   - Create the belief using `mcp__valence_substrate__belief_create`
4. Link the document as a source
5. Report a summary of what was extracted

## Document Path
{{ path }}

{% if domain %}
## Base Domain
{{ domain }}
{% endif %}

## Execution Steps

1. **Read the document**
   - Use the Read tool to get the content
   - Note the document type (markdown, code, config, etc.)

2. **Extract beliefs**
   For each significant piece of information:
   - Formulate it as a clear belief statement
   - Assign confidence:
     - 0.9+ for explicit, unambiguous statements
     - 0.7-0.9 for clear implications
     - 0.5-0.7 for inferences or uncertain claims
   - Classify into domain

3. **Extract entities**
   - People mentioned
   - Tools, libraries, frameworks
   - Projects or codebases
   - Concepts and patterns

4. **Create beliefs**
   Use belief_create for each, linking:
   - The document as source
   - Relevant entities
   - Appropriate domain path

5. **Report results**
   - Number of beliefs extracted
   - Key entities found
   - Any uncertainties or items needing review

## Quality Guidelines

- Don't over-extract: focus on significant, reusable knowledge
- Preserve context: beliefs should be understandable standalone
- Link entities: this enables cross-referencing later
- Note uncertainty: lower confidence for inferred information
- Avoid duplicates: check if similar beliefs already exist

---
> Converted and distributed by [TomeVault](https://tomevault.io/claim/ourochronos) — claim your Tome and manage your conversions.
<!-- tomevault:4.0:skill_md:2026-04-13 -->