/knowledge - View Accumulated Knowledge

Display the current state of accumulated knowledge in the Tree of Skills.

Usage

/knowledge [category]

Categories

• all (default) - Show overview of all knowledge
• literature - Show literature review findings
• domain - Show domain knowledge
• learned - Show experiment learnings
• benchmarks - Show benchmark information
• status - Show project status and phases

Examples

/knowledge                  # Show all knowledge overview
/knowledge literature       # Deep dive into literature
/knowledge benchmarks       # Show benchmark details
/knowledge status           # Show project progress

Process

When user invokes /knowledge, follow this process:

For /knowledge or /knowledge all

Read and display summaries from all skill categories:

╔══════════════════════════════════════════════════════════════╗
║                    ORION KNOWLEDGE BASE                       ║
╠══════════════════════════════════════════════════════════════╣
║ Project: <project-name>                                       ║
║ Status: <current-phase>                                       ║
╚══════════════════════════════════════════════════════════════╝

📚 LITERATURE (X papers analyzed)
├── Methodologies: [count] extracted
├── Experiments: [count] documented
└── Metrics: [count] identified

🧠 DOMAIN KNOWLEDGE
├── Expertise: [summary]
└── Constraints: [summary]

🔬 LEARNED FROM EXPERIMENTS
├── Hypotheses tested: [count]
├── Passed: [count]
└── Key patterns: [summary]

📊 BENCHMARKS
├── Available: [list]
├── Downloaded: [list]
└── Evaluated: [list]

Run /knowledge <category> for details.

For /knowledge literature

Read research/skills/literature/_overview.md and all paper-specific files:

📚 LITERATURE KNOWLEDGE

## Papers Analyzed
| ID | Title | Key Method |
|----|-------|------------|
| P001 | [Title] | [Method] |
| P002 | [Title] | [Method] |

## Methodologies
1. [Method 1]: [Description]
2. [Method 2]: [Description]

## Experiments & Benchmarks (from papers)
- [Paper] used [Benchmark] for [task]
- [Paper] evaluated on [Benchmark]

## Metrics
- [Metric 1]: [What it measures]
- [Metric 2]: [What it measures]

## Key Insights
- [Insight 1]
- [Insight 2]

Deep-dive available: /knowledge literature <paper-id>

For /knowledge domain

Read research/skills/domain/_overview.md and domain skill files:

🧠 DOMAIN KNOWLEDGE

## User Background
- Domain: [domain]
- Expertise Level: [level]
- Specific Areas: [areas]

## Constraints
- Compute: [details]
- Time: [details]
- Data: [details]

## Domain Intuitions
Ideas to explore:
1. [Intuition 1]
2. [Intuition 2]

## Custom Skills
- [Skill 1]: [brief description]
- [Skill 2]: [brief description]

For /knowledge learned

Read research/skills/learned/_overview.md and hypothesis learnings:

🔬 LEARNED FROM EXPERIMENTS

## Hypothesis Summary
| ID | Title | Result | Key Insight |
|----|-------|--------|-------------|
| H-001 | [Title] | PASS/FAIL | [Insight] |
| H-002 | [Title] | PASS/FAIL | [Insight] |

## Patterns Discovered

### What Works ✓
- [Pattern 1]
- [Pattern 2]

### What Doesn't Work ✗
- [Pattern 1] - Reason: [why]
- [Pattern 2] - Reason: [why]

## Best Result So Far
- Hypothesis: [ID]
- Score: [score]
- Delta from baseline: [+X%]

Deep-dive: /knowledge learned <hypothesis-id>

For /knowledge benchmarks

Read research/skills/benchmarks/_overview.md:

📊 BENCHMARK KNOWLEDGE

## Available Benchmarks
| Name | Task | Metrics | Status |
|------|------|---------|--------|
| [Name] | [Task] | [Metrics] | Downloaded/Pending |

## Evaluation Protocols
### [Benchmark 1]
- Task: [description]
- Input format: [format]
- Output format: [format]
- Metrics: [list]
- Subset size: [for quick tests]

## Baseline Results
| Benchmark | Method | Score |
|-----------|--------|-------|
| [Name] | [Method] | [Score] |

For /knowledge status

Read research/orion.json and display project status:

📋 PROJECT STATUS

Project: <project-name>
Current Phase: <phase>

## Phase Progress
✓ Init              - Complete
✓ Literature Review - Complete (X papers)
◯ Benchmark Setup   - In Progress
◯ Hypothesis Gen    - Pending
◯ Experimentation   - Pending

## Key Numbers
- Papers analyzed: X
- Skills extracted: Y
- Benchmarks ready: Z
- Hypotheses: A pending, B tested

## Next Action
[What to do next based on current phase]

Implementation Notes

1. Read files, don't blow up context: Only read _overview.md files unless user asks for deep-dive
2. Format for readability: Use tables and structured output
3. Show actionable next steps: Always suggest what to do next
4. Handle missing data gracefully: If no data in category, show "Not yet populated - run [command]"

Tools Used

• Read: Load skill files
• Direct output: Display formatted knowledge

Output

This command outputs directly to the user - no files are modified. It's a read-only view into the knowledge base.