Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Export a workspace as a portable template. Captures SYSTEM.md, agent definitions, skills, reference files, configuration, seed tasks, and budget defaults. Handles secret stripping and collision metadata for clean import elsewhere. Triggers on: "export", "save template", "package workspace", "share workspace"
.claude/skills/miosa-osa-export/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-15 | ✗→✓ | ▲ Improved | 29% | 0% |
| case-01 | ✗→✓ | ▲ Improved | -9% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 881% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -22% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -56% | 0% |
> Export workspace as a portable, shareable template.
Package a running workspace into a reusable template that can be imported elsewhere. Captures the full workspace definition — agents, skills, configuration, seed tasks, budget defaults — while stripping secrets and environment-specific paths. The exported template is self-contained and portable: anyone with /import can stand up an identical workspace.
bash# Export current workspace /export # Export to specific file /export --output templates/my-research-team.yaml # Export with custom name /export --name "research-team-v2" # Export only structure (no tasks, no data) /export --structure-only # Export including sample data /export --include-samples # Preview what would be exported /export --dry-run
| Flag | Type | Default | Description | |------|------|---------|-------------| | --output | path | ./exported-{name}-{date}.yaml | Output file path | | --name | string | workspace name | Template name | | --description | string | — | Template description | | --structure-only | flag | false | Export structure without tasks or data | | --include-samples | flag | false | Include sample data for reference | | --strip-secrets | flag | true | Remove API keys, tokens, credentials | | --dry-run | flag | false | Preview export without writing | | --format | enum | yaml | yaml, toml, json |
--include-samples).{{SECRET_NAME}} placeholders. Log which secrets need to be provided on import.{{workspace_root}} variables. Replace project-specific names with {{project_name}}.--structure-only, include seed tasks with priorities and assignments.yaml# Workspace Template: research-team-v2 # Exported: 2026-03-20 # Source: /workspace/ai-masters-research name: research-team-v2 description: "3-agent research team for competitive analysis" version: "1.0" exported_at: 2026-03-20T14:30:00Z variables: project_name: description: "Name of the project" default: "Research Project" budget: description: "Total token budget" default: 200000 secrets_required: - ANTHROPIC_API_KEY - OPENAI_API_KEY # for Codex adapter structure: - agents/ - inbox/ - processed/ - ops/observations/ - ops/metrics/ - ops/budget/ agents: - name: researcher role: "Research Analyst" adapter: claude-sonnet budget: "{{budget * 0.4}}" capabilities: [search, read, reduce, reflect] system_prompt: | You are a research analyst focused on {{project_name}}. ... - name: analyst role: "Data Analyst" adapter: claude-sonnet budget: "{{budget * 0.3}}" capabilities: [read, search, graph, stats] - name: writer role: "Content Writer" adapter: claude-haiku budget: "{{budget * 0.15}}" capabilities: [write, edit, translate] skills: - processing/reduce - processing/reflect - analysis/graph - analysis/stats - learning/remember seed_tasks: - task: "Research {{project_name}} competitive landscape" assign: researcher priority: high - task: "Analyze existing data" assign: analyst priority: high after: [0] budget: total: "{{budget}}" alert_threshold: 0.8 hard_ceiling: 1.0
Exported: research-team-v2 → templates/my-research-team.yaml
Agents: 3
Skills: 5
Seed tasks: 2
Variables: 2 (project_name, budget)
Secrets required: 2 (ANTHROPIC_API_KEY, OPENAI_API_KEY)
Size: 4.2 KB/launch — Exported templates are consumed by launch/import — Sister skill for importing| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-15 | fail→pass | 10,373 | 7,783 | -25% | 1 | 1 | 0% | 1,766 | 2,274 | +29% | 0 | 0 | — |
case-01 | fail→pass | 13,960 | 6,491 | -54% | 1 | 1 | 0% | 2,699 | 2,456 | -9% | 0 | 0 | — |
case-02 | fail→pass | 6,890 | 11,950 | +73% | 1 | 1 | 0% | 355 | 3,484 | +881% | 0 | 0 | — |
case-03 | fail→pass | 21,742 | 8,486 | -61% | 1 | 1 | 0% | 3,723 | 2,898 | -22% | 0 | 0 | — |
case-04 | fail→pass | 24,788 | 2,810 | -89% | 1 | 1 | 0% | 4,095 | 1,797 | -56% | 0 | 0 | — |
case-05 | fail→pass | 11,874 | 6,812 | -43% | 1 | 1 | 0% | 1,938 | 2,478 | +28% | 0 | 0 | — |
case-06 | pass→pass | 25,031 | 12,473 | -50% | 1 | 1 | 0% | 2,098 | 3,506 | +67% | 0 | 0 | — |
case-07 | pass→pass | 45,729 | 2,407 | -95% | 1 | 1 | 0% | 3,888 | 1,782 | -54% | 0 | 0 | — |
case-08 | fail→pass | 7,486 | 2,802 | -63% | 1 | 1 | 0% | 1,304 | 1,810 | +39% | 0 | 0 | — |
case-09 | pass→pass | 21,642 | 5,857 | -73% | 1 | 1 | 0% | 2,232 | 2,403 | +8% | 0 | 0 | — |
case-10 | fail→pass | 12,116 | 3,200 | -74% | 1 | 1 | 0% | 2,015 | 1,770 | -12% | 0 | 0 | — |
case-11 | fail→pass | 15,083 | 4,900 | -68% | 1 | 1 | 0% | 2,420 | 2,074 | -14% | 0 | 0 | — |
case-12 | fail→pass | 12,133 | 3,620 | -70% | 1 | 1 | 0% | 1,651 | 1,783 | +8% | 0 | 0 | — |
case-13 | fail→pass | 13,253 | 5,685 | -57% | 1 | 1 | 0% | 2,247 | 1,971 | -12% | 0 | 0 | — |
case-14 | pass→pass | 11,821 | 2,757 | -77% | 1 | 1 | 0% | 1,493 | 1,799 | +20% | 0 | 0 | — |
case-16 | fail→pass | 14,442 | 5,657 | -61% | 1 | 1 | 0% | 1,959 | 2,171 | +11% | 0 | 0 | — |
case-17 | fail→fail | 14,497 | 8,095 | -44% | 1 | 1 | 0% | 1,855 | 2,240 | +21% | 0 | 0 | — |
case-18 | fail→pass | 32,770 | 2,227 | -93% | 1 | 1 | 0% | 1,951 | 1,646 | -16% | 0 | 0 | — |
case-19 | fail→pass | 10,129 | 2,638 | -74% | 1 | 1 | 0% | 1,692 | 1,564 | -8% | 0 | 0 | — |
case-20 | pass→pass | 12,743 | 2,114 | -83% | 1 | 1 | 0% | 2,080 | 1,649 | -21% | 0 | 0 | — |
case-21 | pass→pass | 11,946 | 5,654 | -53% | 1 | 1 | 0% | 1,674 | 2,069 | +24% | 0 | 0 | — |
case-22 | pass→pass | 11,241 | 4,080 | -64% | 1 | 1 | 0% | 1,622 | 1,934 | +19% | 0 | 0 | — |
case-23 | pass→pass | 39,042 | 4,595 | -88% | 1 | 1 | 0% | 4,598 | 1,706 | -63% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted, and 22 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +61 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.