This directory contains standard, executable step-by-step procedures (Workflows) designed to orchestrate agent operations across your projects. Workflows enforce systematic research, data integrity, and strict system maintenance.
| Slash Command / ID | Target Workflow File | Primary Purpose & Usage | Expected Output |
|---|---|---|---|
/research-policy-vn |
research-policy-vn-wf.md | Vietnam Policy Research: Quantitative and sequential multi-agent research into VN's tourism and inbound policies. | Policy Brief report saved to Obsidian Vault Mac/20_Areas/Public_Policy/Tourism_Research/ |
/maintenance-clean |
maintenance-clean-wf.md | Project Maintenance: Pruning obsolete session cache folders (>7 days) and auto-updating standard CHANGELOG/README docs. |
Disk space freed & Git Pulse commit |
/maintenance-vault |
maintenance-vault-wf.md | Obsidian Note Standardizer: Scanning and auto-updating note YAML frontmatter schemas under @Obsidian Vault Mac. |
Standardized notes in Second Brain vault |
/atomic-complexity |
atomic-complexity-wf.md | Modular Decomposition: Auto-splitting complex rules, skills, or tasks exceeding cognitive thresholds into smaller modules. | Structured modularized plan and split files |
/global-flow |
global-flow-wf.md | Meta-Engineering: Unified workflow for improving Antigravity's own rules, skills, and flow optimizations. | Core update and sync |
To run any of the workflows listed above, mention the workflow slash command or reference the file using its file URI or relative workspace link. For example: - "I need to find a tool like HotLingo. Run the /research-oss workflow to research open source alternatives." - "Clean up my old session cache files and sync project docs using the /maintenance-clean workflow." - "Scan my Obsidian Vault and ensure all notes have correct schemas using /maintenance-vault."
The agent will read the corresponding markdown file and execute the steps sequentially, prompting you for any required decisions.
The following specialized agents are defined for orchestrating the Vietnam Tourism Policy Research workspace:
| Agent Identifier / Name | Role Title | Primary Specialty | Target Skills |
|---|---|---|---|
reporting_agent |
Master Policy Conductor | Overseeing the multi-agent flow, interacting with the user, and writing the final master briefs. | grill-me-alignment, ubiquitous-language |
problem_structuring_agent |
Problem Analyst Agent | Structuring policy issues, boundary analysis, identifying systemic failures (market vs. state). | boundary-analysis |
lit_review_agent |
Finder Agent | Coordinating literature search sub-agents (gov_specialist, media_specialist, social_specialist). |
ecosystem-causal-mapping, literature-search-archive |
detail_extractor_agent |
RAG Knowledge Retriever | Running NotebookLM queries via nlm CLI to retrieve qualitative/quantitative evidence. |
nlm-skill |
quantitative_data_agent |
Data Agent | Collecting raw statistics, expenditure patterns, length of stay, and tourism growth indicators. | Quantitative data collection |
research_writer_agent |
Writer Agent | Synthesizing reviews, drafting alternatives, compiling trade-off matrices, CBA/CEA, and M&E. | policy-impact-analysis, backward-mapping |
supervisor_agent |
Reviewer Agent | Fact-checking, citation validation (APA 7th + URLs), and independent NotebookLM checking. | policy-impact-analysis, nlm-skill |
logic_auditor_agent |
Logical & Causal Auditor | Auditing causal logic and reasoning, detecting logical fallacies, and ensuring evidence-backed policy. | Logical and causal auditing |