deep-dive
starsforks
How to install deep-dive?
jmstar85/oh-my-githubcopilot/deep-dive
$npx -y skills add jmstar85/oh-my-githubcopilot --skill deep-dive2-stage pipeline: trace (causal investigation) then deep-interview (requirements). Activate when user says: deep dive, deep-dive, investigate deeply, trace and interview.
SKILL.md
| 1 | # Deep Dive |
| 2 | |
| 3 | Orchestrates a 2-stage pipeline: first investigate WHY something happened (trace), then define WHAT to do about it (deep-interview). Trace findings feed into the interview via 3-point injection. |
| 4 | |
| 5 | ## Pipeline |
| 6 | `deep-dive` → `ralplan` (consensus refinement) → `omg-autopilot` (execution) |
| 7 | |
| 8 | ## When to Use |
| 9 | - User has a problem but doesn't know the root cause |
| 10 | - Bug investigation: "Something broke and I need to figure out why" |
| 11 | - Feature exploration: "I want to improve X but first need to understand it" |
| 12 | |
| 13 | ## When NOT to Use |
| 14 | - Already know the root cause → use `/deep-interview` |
| 15 | - Clear specific request → execute directly |
| 16 | - Investigation only, no requirements → use `/trace` |
| 17 | |
| 18 | ## Phases |
| 19 | |
| 20 | ### Phase 1: Initialize |
| 21 | 1. Parse problem, detect brownfield/greenfield |
| 22 | 2. Generate 3 trace lane hypotheses (code-path, config/env, measurement/artifact) |
| 23 | |
| 24 | ### Phase 2: Lane Confirmation |
| 25 | Present hypotheses to user for confirmation (1 round). |
| 26 | |
| 27 | ### Phase 3: Trace Execution |
| 28 | Run 3 parallel tracer lanes using @tracer agents: |
| 29 | - Each lane: evidence for, evidence against, critical unknown, discriminating probe |
| 30 | - Rebuttal round between top hypotheses |
| 31 | - Convergence detection |
| 32 | - Save to `.omg/specs/deep-dive-trace-{slug}.md` |
| 33 | |
| 34 | ### Phase 4: Interview with Trace Injection |
| 35 | Follow deep-interview protocol with 3 overrides: |
| 36 | 1. **initial_idea enrichment**: Include trace's most likely explanation |
| 37 | 2. **codebase_context replacement**: Use trace synthesis (skip re-exploration) |
| 38 | 3. **question queue injection**: Per-lane critical unknowns become first questions |
| 39 | |
| 40 | Low-confidence trace: don't inject uncertain conclusion, use ALL unknowns as questions. |
| 41 | |
| 42 | ### Phase 5: Execution Bridge |
| 43 | Same options as deep-interview: ralplan → omg-autopilot (recommended), omg-autopilot, ralph, team, or refine further. |
| 44 | |
| 45 | ## Output |
| 46 | Spec saved to `.omg/specs/deep-dive-{slug}.md` with additional "Trace Findings" section. |
People who liked this also viewed
6 picksType
- find-skillsHelps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skSkillsJul 2026
- grill-meA relentless interview to sharpen a plan or design.SkillsJul 2026
- vercel-react-best-practicesVercel's official collection of agent skillsSkillsJul 2026
- improve-codebase-architectureScan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.SkillsJul 2026
- setup-matt-pocock-skillsConfigure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.SkillsJul 2026
- triageMove issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.SkillsJul 2026
