with one click
investigation
Deep-dive investigation with documentation output.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Menu
Deep-dive investigation with documentation output.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
| name | investigation |
| description | Deep-dive investigation with documentation output. |
Conducts deep-dive investigations on specific topics and produces detailed documentation.
Session path: [SESSION_PATH from working memory] Topic: $ARGUMENTS
Investigate thoroughly:
Create documentation:
[SESSION_PATH]/[TIMESTAMP_FILE]-dive-[topic-slug].mdStructure:
# Deep Dive: [topic]
Date: [TIMESTAMP_LOG]
## Summary
[Brief overview of findings]
## Key Findings
[Bullet points]
## Code Structure
[Relevant files with brief explanations]
## Dependencies & Relationships
[How components interact]
## Considerations
[Issues, edge cases, concerns]
## Recommendations
[Suggested next steps]
Update session:
[SESSION_PATH]/_overview.md:
- [TIMESTAMP_ITERATION] Deep dive: [topic] -> [filename].md- [filename].md - Deep dive: [topic]cd [SESSION_DIR] && git add . && git commit -m "Deep dive: [topic]"Report back: Provide concise summary of key findings with filename
Investigate GitHub pull request review comments, determine whether each comment is valid, propose fixes, and document the analysis in an active samocode session. Use when the user asks to analyze PR comments, run /prcomments, investigate review feedback, or triage pull request comments.
Run OpenAI Codex CLI as a subagent for second opinions, code reviews, and questions. Use when you want a different AI model's perspective.
Run Anthropic Claude CLI as a subagent for second opinions, code reviews, and questions. Use from a Codex (or other) session when you want Claude's perspective.
Create GitHub PR review comments from review findings. Use after running a code review to post findings as line-bound comments.
Execute implementation tasks with different approaches (single, dual-agent, plan-based).
Create implementation plans with phase management.