Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Migrate an existing Claude, Copilot, Codex, Cursor, Kiro, VS Code, or other client-specific agent plugin to the portable Agent Plugins v1 structure while preserving platform-specific hooks, agents, commands, LSP, UI, and marketplace behavior. Use when auditing, converting, or modernizing an agent plugin.
.claude/skills/vinhnx-migrate-agent-plugin/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -8% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -5% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -14% | 0% |
Convert an existing plugin to the Agent Plugins v1 portable core without prematurely removing behavior required by its current clients.
Use the current Agent Plugins specification as the normative source.
Read these references before editing:
plugin.json, Agent Skills in skills/, and MCP servers in mcp.json.plugin.json at the plugin root.$schema to https://agent-plugins.org/schemas/1.0.0/plugin.schema.json.name and only supported metadata fields.hooks, agents, skills, or mcpServers at the top level.skills/<skill-name>/SKILL.md; only immediate children of skills/ are discovered.mcp.json, declare the matching v1.0.0 schema, and give every server an explicit stdio, streamable-http, or sse type.${PLUGIN_ROOT} for packaged read-only resources and ${PLUGIN_DATA} for persistent writable state where the MCP schema permits expansion.Before finishing, report:
Prefer an additive, reversible migration. Never claim that hooks, agents, commands, LSP servers, UI, or marketplace metadata became portable Agent Plugins v1 components.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-05 | pass→pass | 21,350 | 24,877 | +17% | 1 | 1 | 0% | 3,256 | 4,716 | +45% | 0 | 0 | — |
case-01 | fail→fail | 21,946 | 13,929 | -37% | 1 | 1 | 0% | 4,397 | 989 | -78% | 0 | 0 | — |
case-02 | fail→fail | 26,129 | 12,662 | -52% | 1 | 1 | 0% | 5,795 | 991 | -83% | 0 | 0 | — |
case-03 | fail→fail | 28,646 | 11,090 | -61% | 1 | 1 | 0% | 4,715 | 1,023 | -78% | 0 | 0 | — |
case-04 | pass→pass | 16,831 | 16,477 | -2% | 1 | 1 | 0% | 3,185 | 3,017 | -5% | 0 | 0 | — |
case-06 | pass→pass | 25,034 | 16,630 | -34% | 1 | 1 | 0% | 4,767 | 4,371 | -8% | 0 | 0 | — |
case-07 | fail→pass | 17,377 | 15,562 | -10% | 1 | 1 | 0% | 2,297 | 2,927 | +27% | 0 | 0 | — |
case-08 | fail→pass | 18,996 | 8,002 | -58% | 1 | 1 | 0% | 2,529 | 2,321 | -8% | 0 | 0 | — |
case-09 | fail→pass | 15,544 | 11,067 | -29% | 1 | 1 | 0% | 1,841 | 1,972 | +7% | 0 | 0 | — |
case-10 | fail→pass | 16,949 | 4,392 | -74% | 1 | 1 | 0% | 1,781 | 1,684 | -5% | 0 | 0 | — |
case-11 | fail→pass | 17,301 | 10,985 | -37% | 1 | 1 | 0% | 2,311 | 1,992 | -14% | 0 | 0 | — |
case-12 | fail→pass | 19,421 | 15,756 | -19% | 1 | 1 | 0% | 2,647 | 2,705 | +2% | 0 | 0 | — |
case-13 | fail→pass | 9,819 | 11,095 | +13% | 1 | 1 | 0% | 1,716 | 1,938 | +13% | 0 | 0 | — |
case-14 | fail→pass | 18,437 | 5,694 | -69% | 1 | 1 | 0% | 2,870 | 1,642 | -43% | 0 | 0 | — |
case-15 | fail→pass | 21,184 | 15,049 | -29% | 1 | 1 | 0% | 2,770 | 2,534 | -9% | 0 | 0 | — |
case-16 | pass→pass | 14,164 | 9,210 | -35% | 1 | 1 | 0% | 1,526 | 1,428 | -6% | 0 | 0 | — |
case-17 | pass→pass | 10,750 | 8,384 | -22% | 1 | 1 | 0% | 1,742 | 2,023 | +16% | 0 | 0 | — |
case-18 | pass→pass | 19,153 | 13,743 | -28% | 1 | 1 | 0% | 2,410 | 2,041 | -15% | 0 | 0 | — |
case-19 | fail→pass | 21,568 | 15,720 | -27% | 1 | 1 | 0% | 3,076 | 3,130 | +2% | 0 | 0 | — |
case-20 | fail→pass | 15,740 | 12,348 | -22% | 1 | 1 | 0% | 1,821 | 1,993 | +9% | 0 | 0 | — |
case-21 | fail→pass | 23,074 | 15,967 | -31% | 1 | 1 | 0% | 2,457 | 2,292 | -7% | 0 | 0 | — |
case-22 | fail→pass | 12,292 | 9,088 | -26% | 1 | 1 | 0% | 1,676 | 1,516 | -10% | 0 | 0 | — |
case-23 | pass→pass | 15,324 | 10,722 | -30% | 1 | 1 | 0% | 1,655 | 1,755 | +6% | 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 20 counted toward the lift figure. The other 3 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 +57 percentage points is the difference between those two pass rates over the 20 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
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.