- SSOT:
.agents/directory (do not modify directly) - Response language: Follows
languagein.agents/oma-config.yaml - Skills:
.agents/skills/(domain specialists) - Workflows:
.agents/workflows/(multi-step orchestration) - Subagents: Same-vendor native dispatch via
.gemini/agents/{name}.md; cross-vendor or unsupported cases fall back tooma agent:spawn {agent} {prompt} {sessionId}
- Resolve
target_vendor_for_agentfrom.agents/oma-config.yaml. - If
target_vendor_for_agent === current_runtime_vendor, use the runtime's native subagent path. - If vendors differ, or native subagents are unavailable, use
oma agent:spawnfor that agent only.
Prefer serena MCP tools over native find/grep when locating code — they are symbol-aware and faster on large repos. Fall back to native Read / Glob / Grep only when serena is unavailable or for plain file content reads.
| Task | Preferred tool |
|---|---|
| Locate a symbol definition (class / function / variable) | find_symbol |
| Find references / callers of a symbol | find_referencing_symbols |
| Outline a file's top-level symbols | get_symbols_overview |
| Pattern or regex search across the codebase | search_for_pattern |
| Find a file by name | find_file |
| List directory contents | list_dir |
Execute by naming the workflow in your prompt. Keywords are auto-detected via hooks.
| Workflow | File | Description |
|---|---|---|
| orchestrate | orchestrate.md |
Parallel subagents + Review Loop |
| work | work.md |
Step-by-step with remediation loop |
| ultrawork | ultrawork.md |
5-Phase Gate Loop (11 reviews) |
| plan | plan.md |
PM task breakdown |
| brainstorm | brainstorm.md |
Design-first ideation |
| review | review.md |
QA audit |
| debug | debug.md |
Root cause + minimal fix |
| deepsec | deepsec.md |
Drive oma-deepsec end-to-end (setup / scan / pr-review / matchers / triage) |
| scm | scm.md |
SCM + Git operations + Conventional Commits |
| docs | docs.md |
Documentation drift verify + sync |
| recap | recap.md |
Daily / period AI conversation recap |
To execute: read and follow .agents/workflows/{name}.md step by step.
Hooks: BeforeAgent (keyword detection), BeforeTool, AfterAgent (persistent mode)
Keywords defined in .agents/hooks/core/triggers.json (multi-language).
Persistent workflows (orchestrate, ultrawork, work) block termination until complete.
Deactivate: say "workflow done".
- Do not modify
.agents/files (SSOT protection). - Workflows execute via keyword detection or explicit naming, never self-initiated.
- Response language follows
.agents/oma-config.yaml
Read the relevant file from .agents/rules/ when working on matching code.
| Rule | File | Scope |
|---|---|---|
| backend | .agents/rules/backend.md |
on request |
| commit | .agents/rules/commit.md |
on request |
| database | .agents/rules/database.md |
**/*.{sql,prisma} |
| debug | .agents/rules/debug.md |
on request |
| design | .agents/rules/design.md |
on request |
| dev-workflow | .agents/rules/dev-workflow.md |
on request |
| frontend | .agents/rules/frontend.md |
**/*.{tsx,jsx,css,scss} |
| i18n-guide | .agents/rules/i18n-guide.md |
always |
| infrastructure | .agents/rules/infrastructure.md |
**/*.{tf,tfvars,hcl} |
| market | .agents/rules/market.md |
on request |
| mobile | .agents/rules/mobile.md |
**/*.{dart,swift,kt} |
| quality | .agents/rules/quality.md |
on request |