/deep-dive
Use this skill when requirements are incomplete, conflicting, or too vague to begin implementation safely.
$ npx -y skills add Joonghyun-Lee-Frieren/oh-my-antigravity --skill deep-dive --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
- Slash command
/deep-dive
Context preview
The summary Claude sees to decide when to auto-load this skill.
Use this skill when requirements are incomplete, conflicting, or too vague to begin implementation safely.
SKILL.md
deep-dive.SKILL.mdname = "deep-dive"
description = "Run a trace-to-interview discovery pass to remove ambiguity before planning and execution."
Purpose
Use this skill when requirements are incomplete, conflicting, or too vague to begin implementation safely.
Trigger
- User asks for implementation but goals/constraints are still unclear
- Multiple interpretations exist for scope, acceptance criteria, or architecture direction
- Team needs a compact interview artifact before `/oma:team-plan` or `/oma:team-prd`
Workflow
1. Trace known facts from the request and current repository context. 2. Compute a simple clarity score (`low`, `medium`, `high`) for implementation readiness. 3. Run a structured interview pass:
- essential questions (must answer before planning)
- standard questions (quality/cost/risk tuning)
- deep questions (edge-case and long-term maintainability)
4. Record assumptions explicitly and mark each as validated, pending, or risky. 5. Produce a launch brief that can be consumed by planning/execution stages. 6. If filesystem tools are available, update:
- `.omg/state/interview-context.json`
- `.omg/state/launch-brief.md`
Output Template
## Clarity Score
- level: low|medium|high
- rationale: ...
## Known Facts
- ...
## Open Questions
1. ...
2. ...
## Assumption Ledger
| Assumption | Status (validated/pending/risky) | Validation Path |
| --- | --- | --- |
| ... | ... | ... |
## Launch Brief
- objective:
- boundaries:
- acceptance hints:
- recommended next command:
Notes
- Keep output concise and operator-facing.
- Do not start coding when essential interview questions are unresolved.
- This skill remains extension-native (prompt/state level), without external daemons or runtime wrappers.
Read more
name = "deep-dive" description = "Run a trace-to-interview discovery pass to remove ambiguity before planning and execution."
Purpose
Use this skill when requirements are incomplete, conflicting, or too vague to begin implementation safely.
Trigger
- User asks for implementation but goals/constraints are still unclear
- Multiple interpretations exist for scope, acceptance criteria, or architecture direction
- Team needs a compact interview artifact before `/oma:team-plan` or `/oma:team-prd`
Workflow
1. Trace known facts from the request and current repository context. 2. Compute a simple clarity score (`low`, `medium`, `high`) for implementation readiness. 3. Run a structured interview pass:
- essential questions (must answer before planning)
- standard questions (quality/cost/risk tuning)
- deep questions (edge-case and long-term maintainability)
4. Record assumptions explicitly and mark each as validated, pending, or risky. 5. Produce a launch brief that can be consumed by planning/execution stages. 6. If filesystem tools are available, update:
- `.omg/state/interview-context.json`
- `.omg/state/launch-brief.md`
Output Template
## Clarity Score - level: low|medium|high - rationale: ... ## Known Facts - ... ## Open Questions 1. ... 2. ... ## Assumption Ledger | Assumption | Status (validated/pending/risky) | Validation Path | | --- | --- | --- | | ... | ... | ... | ## Launch Brief - objective: - boundaries: - acceptance hints: - recommended next command:
Notes
- Keep output concise and operator-facing.
- Do not start coding when essential interview questions are unresolved.
- This skill remains extension-native (prompt/state level), without external daemons or runtime wrappers.
Antigravity CLI: Gemini CLI will be replaced by Antigravity CLI on June 18th. If you previously installed oh-my-gemini-cli in Gemini CLI, Antigravity CLI will ask whether to import plugins during installation. Check that option to bring your plugin over.
Repo: Joonghyun-Lee-Frieren/oh-my-antigravity
Other skills on oh-my-antigravity.
- /blueprint
Use this skill when a product, UI, or workflow decision needs a stable decision surface before implementation.
Open skill - /context-optimize
Use this skill when context is growing quickly or cache consistency drops.
Open skill - /execute
Use this skill when the task is implementation-ready and requires code changes now.
Open skill - /learn
Automatically evaluates OmA sessions to extract reusable patterns (error resolutions, workarounds, conventions) and save them to `.omg/rules/learned/`.
Open skill - /oma-plan
Use this skill when the request is non-trivial and should be planned before editing code.
Open skill - /plan
Use this skill when the request is non-trivial and should be planned before editing code.
Open skill

