Oculus Proficiency Test

This test validates that an agent can effectively use Oculus for navigation, reading, and editing based on the CLAUDE.md guidance.

Test Objective

Complete all tasks in the checklist below. Each task exercises a different aspect of Oculus usage.

Tasks

Tasks

Tasks [proficiency-test-tasks]

• Task 1: Find and Read - Use _____grep to search for nodes containing "tagging-taxonomy". View the node using ______look and extract the tag format convention.
• Task 2: Explore Structure (Index Format) - The cmd-poke node is large (~26KB). Use ______look({ slug: "cmd-poke", format: "index" }) to explore its structure without loading full content. Document: (a) how many sections it has, (b) how many fences it contains, and (c) the index number of ANY fence you find interesting.
• Task 3: Extract Data (Peek) - Now use _____peek to extract a specific value from the cmd-poke node. Use the structure you discovered in Task 2 to choose a path. Document what path you used and what value you found.
• Task 4: Create Documentation - Create a new node called test-agent-notes-<identifier> where <identifier> is your session ID, name, or timestamp (e.g., test-agent-notes-claude, test-agent-notes-20260122b, or test-agent-notes-session-abc). Add a summary of what you learned about (a) tagging conventions and (b) index format benefits.
• Task 5: Link Your Work - Link your new notes node to this test node (use south direction from this node, backlink=true).
• Task 6: Add Section - Add a new H2 section to your notes node called "## Test Results". First use ______look to find your H1 slug, then poke to that H1 slug with operation: "append" and include the "## Test Results" header in your value.
• Task 7: Complete Test - Mark all tasks above as complete using the virtual fence execute action.

Expected Skills

• Navigation: Using goto, look, grep, and search tools
• Context Management: Using enhanced look formats (index) to explore large nodes efficiently
• Reading: Using peek with proper path syntax
• Writing: Using poke with correct operation modes
• Graph Management: Creating nodes, linking them, tagging them
• Todo Management: Executing virtual fence actions on todo lists

Success Criteria

All checkboxes above are checked, and the test-agent-notes node exists with:

• Summary of tagging conventions (from Task 1)
• Summary of index format benefits (from Task 2)
• Documentation of structure exploration findings (section count, fence count, fence index)
• Documentation of peek extraction (from Task 3)
• A "Test Results" section (from Task 6)
• A south link from this test node (from Task 5)

Tips

• Review the CLAUDE.md "Oculus Navigation and Editing - How and Why" section
• Remember: to add a new section, poke to the PARENT with the header included
• Always include context parameter when poking
• Use ______look to inspect node structure before editing
• Tip: For large nodes, use ______look({ format: "index" }) to see structure without loading full content

Provenance

Document

• Status: 🔴 Unverified

Changelog

• 2026-01-22 15:57: Node created by mcp - Creating test task to verify CLAUDE.md Oculus guidance

East

slots:
- slug: oculus-proficiency-test-questionnaire
  context:
  - Linking questionnaire template to main test for reference
- slug: oculus-proficiency-test-20260123-a-grading
  context:
  - Linking grading report to test instance

South

slots:
- slug: test-sandbox
  context:
  - Linking sandbox to test for bonus task access
- slug: test-agent-notes-opus-20260123
  context:
  - Linking proficiency test notes as completion evidence

Tasks Proficiency Test Tasks