Relativity Deposition Question Builder
name: relativity-deposition-question-builder
by chipmunkrpa · published 2026-03-22
$ claw add gh:chipmunkrpa/chipmunkrpa-deposition-question-skill---
name: relativity-deposition-question-builder
description: Develop deposition question sets from Relativity-exported PDF productions using a user-provided legal theory. Use when tasks involve reviewing opponent-produced PDFs, extracting per-page document IDs from the bottom-right corner (if two numbers appear, use the smaller one), mapping document relevance to a legal theory, and drafting questions grouped by document ID with required rationale and supporting quotes.
---
# Relativity Deposition Question Builder
Overview
Analyze one or more Relativity-exported PDF productions against a legal theory and generate deposition questions organized by document ID. Each question includes a reason for asking and a document quote that can be used if the witness denies.
Required Behavior
- `Reason why we ask this question`
- `Quote from the document to use in deposition in case the opponent party denies`
Workflow
1. Gather inputs.
- Confirm the user's legal theory.
- Confirm PDF source path(s).
- Optionally gather witness name/role and priority topics.
2. Extract per-page text and document IDs.
- Run:
```bash
python scripts/extract_relativity_pages.py \
--input <pdf-folder-or-file> \
--recurse \
--output <relativity_pages.json>
```
- Review pages where `selected_document_id` is null and flag for manual check.
3. Build a relevance map for the legal theory.
- For each page, classify relation to theory: `supports`, `undermines`, `neutral`.
- Focus question drafting on `supports` pages first, then `undermines` pages for impeachment.
4. Draft deposition questions grouped by document ID.
- Use short, concrete, single-issue questions.
- Start with authentication/foundation, then admission and contradiction questions.
- Use direct quotes from page text for denial follow-up.
5. Return output in required structure.
- Follow `references/deposition_output_template.md`.
- Keep document IDs in ascending numeric order.
Output Rules
- `Quote from the document to use in deposition in case the opponent party denies: [No direct quote located - manual verification required]`
Resources
Dependencies
Install once if missing:
python -m pip install --user pdfplumberMore tools from the same signal band
Order food/drinks (点餐) on an Android device paired as an OpenClaw node. Uses in-app menu and cart; add goods, view cart, submit order (demo, no real payment).
Sign plugins, rotate agent credentials without losing identity, and publicly attest to plugin behavior with verifiable claims and authenticated transfers.
The philosophical layer for AI agents. Maps behavior to Spinoza's 48 affects, calculates persistence scores, and generates geometric self-reports. Give your...