Files
everyinc--compound-engineer…/tests/converter.test.ts
T
2026-07-13 12:20:01 +08:00

682 lines
25 KiB
TypeScript

import { describe, expect, test } from "bun:test"
import { promises as fs } from "fs"
import path from "path"
import { loadClaudePlugin } from "../src/parsers/claude"
import { convertClaudeToOpenCode, transformSkillContentForOpenCode } from "../src/converters/claude-to-opencode"
import { parseFrontmatter } from "../src/utils/frontmatter"
import type { ClaudePlugin } from "../src/types/claude"
const fixtureRoot = path.join(import.meta.dir, "fixtures", "sample-plugin")
const compoundEngineeringRoot = path.join(import.meta.dir, "..")
describe("convertClaudeToOpenCode", () => {
test("current compound-engineering output is skills only, no standalone agents, with one command stub per skill", async () => {
const plugin = await loadClaudePlugin(compoundEngineeringRoot)
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: true,
permissions: "none",
})
expect(bundle.agents).toHaveLength(0)
expect(bundle.skillDirs.length).toBeGreaterThan(0)
// Each user-invocable skill now also generates a slash-command stub
// (skills with disable-model-invocation or user-invocable: false are excluded).
expect(bundle.commandFiles.length).toBeGreaterThan(0)
expect(bundle.commandFiles.length).toBeLessThanOrEqual(bundle.skillDirs.length)
expect(bundle.plugins).toHaveLength(0)
expect(bundle.config.tools).toBeUndefined()
})
test("skills generate slash-command stubs with description and argument-hint frontmatter", async () => {
const plugin = await loadClaudePlugin(fixtureRoot)
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "none",
})
const cmd = bundle.commandFiles.find((f) => f.name === "skill-one")
expect(cmd).toBeDefined()
const parsed = parseFrontmatter(cmd!.content)
expect(parsed.data.description).toBe("Sample skill")
expect(parsed.body.trim()).toContain("skill-one")
expect(parsed.body.trim()).toContain("$ARGUMENTS")
// disabled-skill (disable-model-invocation) is user-invocation-only, so it
// still gets a slash command -- that's its only entry point.
expect(bundle.commandFiles.find((f) => f.name === "disabled-skill")).toBeDefined()
// claude-only-skill is platform-filtered before convertSkillsToCommands -- absent
expect(bundle.commandFiles.find((f) => f.name === "claude-only-skill")).toBeUndefined()
// agent-only-skill has user-invocable: false -- must not be exposed as a slash command
expect(bundle.commandFiles.find((f) => f.name === "agent-only-skill")).toBeUndefined()
})
test("no command name is emitted more than once", async () => {
const plugin = await loadClaudePlugin(fixtureRoot)
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "none",
})
const names = bundle.commandFiles.map((f) => f.name)
const uniqueNames = new Set(names)
expect(names.length).toBe(uniqueNames.size)
})
test("explicit command foo:bar blocks skill stub foo/bar from being emitted", async () => {
// "foo:bar" (explicit command) and "foo/bar" (skill name) both normalize to
// the same on-disk path commands/foo/bar.md -- the skill stub must be dropped.
const plugin: ClaudePlugin = {
root: "/tmp/fake-plugin",
manifest: { name: "test-plugin", version: "1.0.0", description: "" },
agents: [],
commands: [{ name: "foo:bar", description: "explicit", body: "explicit body", sourcePath: "/tmp/fake-plugin/commands/foo/bar.md" }],
skills: [
{
name: "foo/bar",
description: "skill",
sourceDir: "/tmp/fake-plugin/skills/foo-bar",
skillPath: "/tmp/fake-plugin/skills/foo-bar/SKILL.md",
},
],
mcpServers: undefined,
hooks: undefined,
}
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "none",
})
const fooBarCommands = bundle.commandFiles.filter((f) =>
f.name === "foo:bar" || f.name === "foo/bar",
)
// Only the explicit command should survive
expect(fooBarCommands).toHaveLength(1)
expect(fooBarCommands[0].name).toBe("foo:bar")
expect(fooBarCommands[0].content).toContain("explicit body")
})
test("from-commands mode: skill tool is allowed when skill stubs are emitted", async () => {
// Regression: applyPermissions only scanned plugin.commands for allowedTools,
// so a plugin with no explicit commands (or none listing "skill") would get
// permission.skill = "deny", causing every generated stub to fail immediately.
const plugin = await loadClaudePlugin(fixtureRoot)
// Build a minimal plugin with skills but no explicit commands.
const skillOnlyPlugin: ClaudePlugin = {
...plugin,
commands: [],
}
const bundle = convertClaudeToOpenCode(skillOnlyPlugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "from-commands",
})
// Skill stubs should be emitted (skill-one is opencode-eligible)
expect(bundle.commandFiles.some((f) => f.name === "skill-one")).toBe(true)
// The skill tool must be allowed so the stubs can actually load the skill
const permission = bundle.config.permission as Record<string, string>
expect(permission.skill).toBe("allow")
})
test("from-commands mode: patterned skill(...) rule does not block other skill stubs", async () => {
// Regression: when a command uses skill(foo-*), patterns["skill"] is populated,
// causing the permission build to emit { "*": "deny", "foo-*": "allow" }.
// That blocks all stubs for skills outside the pattern even though hasSkillStubs
// forces enabled.add("skill"). The fix clears patterns["skill"] when hasSkillStubs is true.
const plugin = await loadClaudePlugin(fixtureRoot)
// The fixture has skill-command.md with allowed-tools: Skill(create-agent-skills).
// skill-one is the stub-eligible skill -- it is NOT in the patterned subset.
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "from-commands",
})
expect(bundle.commandFiles.some((f) => f.name === "skill-one")).toBe(true)
// skill must be a flat "allow", not a pattern object that would deny skill-one
const permission = bundle.config.permission as Record<string, string | Record<string, string>>
expect(permission.skill).toBe("allow")
})
test("from-command mode: map allowedTools to global permission block", async () => {
const plugin = await loadClaudePlugin(fixtureRoot)
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "from-commands",
})
expect(bundle.config.command).toBeUndefined()
expect(bundle.config.tools).toBeUndefined()
expect(bundle.commandFiles.find((f) => f.name === "workflows:review")).toBeDefined()
expect(bundle.commandFiles.find((f) => f.name === "plan_review")).toBeDefined()
const permission = bundle.config.permission as Record<string, string | Record<string, string>>
expect(Object.keys(permission).sort()).toEqual([
"bash",
"edit",
"glob",
"grep",
"list",
"patch",
"question",
"read",
"skill",
"task",
"todoread",
"todowrite",
"webfetch",
"write",
])
expect(permission.edit).toBe("allow")
expect(permission.write).toBe("allow")
const bashPermission = permission.bash as Record<string, string>
expect(bashPermission["ls *"]).toBe("allow")
expect(bashPermission["git *"]).toBe("allow")
expect(permission.webfetch).toBe("allow")
const readPermission = permission.read as Record<string, string>
expect(readPermission["*"]).toBe("deny")
expect(readPermission[".env"]).toBe("allow")
expect(permission.question).toBe("allow")
expect(permission.todowrite).toBe("allow")
expect(permission.todoread).toBe("allow")
const agentFile = bundle.agents.find((agent) => agent.name === "repo-research-analyst")
expect(agentFile).toBeDefined()
const parsed = parseFrontmatter(agentFile!.content)
expect(parsed.data.mode).toBe("subagent")
})
test("normalizes models and infers temperature", async () => {
const plugin = await loadClaudePlugin(fixtureRoot)
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "primary",
inferTemperature: true,
permissions: "none",
})
const securityAgent = bundle.agents.find((agent) => agent.name === "security-sentinel")
expect(securityAgent).toBeDefined()
const parsed = parseFrontmatter(securityAgent!.content)
expect(parsed.data.model).toBe("anthropic/claude-sonnet-4-20250514")
expect(parsed.data.temperature).toBe(0.1)
const modelCommand = bundle.commandFiles.find((f) => f.name === "workflows:work")
expect(modelCommand).toBeDefined()
const commandParsed = parseFrontmatter(modelCommand!.content)
expect(commandParsed.data.model).toBe("openai/gpt-4o")
})
test("resolves bare Claude model aliases for primary agents", () => {
const plugin: ClaudePlugin = {
root: "/tmp/plugin",
manifest: { name: "fixture", version: "1.0.0" },
agents: [
{
name: "cheap-agent",
description: "Agent using bare alias",
body: "Test agent.",
sourcePath: "/tmp/plugin/agents/cheap-agent.md",
model: "haiku",
},
],
commands: [],
skills: [],
}
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "primary",
inferTemperature: false,
permissions: "none",
})
const agent = bundle.agents.find((a) => a.name === "cheap-agent")
expect(agent).toBeDefined()
const parsed = parseFrontmatter(agent!.content)
expect(parsed.data.model).toBe("anthropic/claude-haiku-4-5")
})
test("suppresses inferred temperature for primary agents on sampling-param-rejecting models", () => {
const plugin: ClaudePlugin = {
root: "/tmp/plugin",
manifest: { name: "fixture", version: "1.0.0" },
agents: [
{
name: "security-guardian",
description: "Reviews code for security issues",
body: "Test agent.",
sourcePath: "/tmp/plugin/agents/security-guardian.md",
model: "sonnet",
},
{
name: "haiku-guardian",
description: "Reviews code cheaply",
body: "Test agent.",
sourcePath: "/tmp/plugin/agents/haiku-guardian.md",
model: "haiku",
},
],
commands: [],
skills: [],
}
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "primary",
inferTemperature: true,
permissions: "none",
})
// Sonnet 5 rejects non-default temperature with HTTP 400, so the converter
// must write the model but omit the inferred temperature.
const sonnetAgent = parseFrontmatter(bundle.agents.find((a) => a.name === "security-guardian")!.content)
expect(sonnetAgent.data.model).toBe("anthropic/claude-sonnet-5")
expect(sonnetAgent.data.temperature).toBeUndefined()
// Haiku still accepts sampling params, so temperature is inferred as usual.
const haikuAgent = parseFrontmatter(bundle.agents.find((a) => a.name === "haiku-guardian")!.content)
expect(haikuAgent.data.model).toBe("anthropic/claude-haiku-4-5")
expect(haikuAgent.data.temperature).toBe(0.1)
})
test("omits model for subagents to allow provider inheritance (#477)", () => {
const plugin: ClaudePlugin = {
root: "/tmp/plugin",
manifest: { name: "fixture", version: "1.0.0" },
agents: [
{
name: "cheap-agent",
description: "Agent using bare alias",
body: "Test agent.",
sourcePath: "/tmp/plugin/agents/cheap-agent.md",
model: "haiku",
},
],
commands: [],
skills: [],
}
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "none",
})
const agent = bundle.agents.find((a) => a.name === "cheap-agent")
expect(agent).toBeDefined()
const parsed = parseFrontmatter(agent!.content)
expect(parsed.data.model).toBeUndefined()
})
test("omits model when agent has no model field regardless of mode", () => {
const plugin: ClaudePlugin = {
root: "/tmp/plugin",
manifest: { name: "fixture", version: "1.0.0" },
agents: [
{
name: "no-model-agent",
description: "Agent without model",
body: "Test agent.",
sourcePath: "/tmp/plugin/agents/no-model-agent.md",
},
],
commands: [],
skills: [],
}
for (const mode of ["primary", "subagent"] as const) {
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: mode,
inferTemperature: false,
permissions: "none",
})
const agent = bundle.agents.find((a) => a.name === "no-model-agent")
const parsed = parseFrontmatter(agent!.content)
expect(parsed.data.model).toBeUndefined()
}
})
test("omits model: inherit even in primary mode", () => {
const plugin: ClaudePlugin = {
root: "/tmp/plugin",
manifest: { name: "fixture", version: "1.0.0" },
agents: [
{
name: "inherit-agent",
description: "Agent with inherit model",
body: "Test agent.",
sourcePath: "/tmp/plugin/agents/inherit-agent.md",
model: "inherit",
},
],
commands: [],
skills: [],
}
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "primary",
inferTemperature: false,
permissions: "none",
})
const agent = bundle.agents.find((a) => a.name === "inherit-agent")
const parsed = parseFrontmatter(agent!.content)
expect(parsed.data.model).toBeUndefined()
})
test("converts hooks into plugin file", async () => {
const plugin = await loadClaudePlugin(fixtureRoot)
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "none",
})
const hookFile = bundle.plugins.find((file) => file.name === "converted-hooks.ts")
expect(hookFile).toBeDefined()
expect(hookFile!.content).toContain("\"tool.execute.before\"")
expect(hookFile!.content).toContain("\"tool.execute.after\"")
expect(hookFile!.content).toContain("\"session.created\"")
expect(hookFile!.content).toContain("\"session.deleted\"")
expect(hookFile!.content).toContain("\"session.idle\"")
expect(hookFile!.content).toContain("\"experimental.session.compacting\"")
expect(hookFile!.content).toContain("\"permission.requested\"")
expect(hookFile!.content).toContain("\"permission.replied\"")
expect(hookFile!.content).toContain("\"message.created\"")
expect(hookFile!.content).toContain("\"message.updated\"")
expect(hookFile!.content).toContain("echo before")
expect(hookFile!.content).toContain("echo before two")
expect(hookFile!.content).toContain("// timeout: 30s")
expect(hookFile!.content).toContain("// Prompt hook for Write|Edit")
expect(hookFile!.content).toContain("// Agent hook for Write|Edit: security-sentinel")
// PreToolUse (tool.execute.before) handlers are wrapped in try-catch
// to prevent hook failures from crashing parallel tool call batches (#85)
const beforeIdx = hookFile!.content.indexOf('"tool.execute.before"')
const afterIdx = hookFile!.content.indexOf('"tool.execute.after"')
const beforeBlock = hookFile!.content.slice(beforeIdx, afterIdx)
expect(beforeBlock).toContain("try {")
expect(beforeBlock).toContain("} catch (err) {")
// PostToolUse (tool.execute.after) handlers are NOT wrapped in try-catch
const afterBlock = hookFile!.content.slice(afterIdx, hookFile!.content.indexOf('"session.created"'))
expect(afterBlock).not.toContain("try {")
})
test("converts MCP servers", async () => {
const plugin = await loadClaudePlugin(fixtureRoot)
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "none",
})
const mcp = bundle.config.mcp ?? {}
expect(mcp["local-tooling"]).toEqual({
type: "local",
command: ["echo", "fixture"],
environment: undefined,
enabled: true,
})
expect(mcp.context7).toEqual({
type: "remote",
url: "https://mcp.context7.com/mcp",
headers: undefined,
enabled: true,
})
})
test("permission modes set expected keys", async () => {
const plugin = await loadClaudePlugin(fixtureRoot)
const noneBundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "none",
})
expect(noneBundle.config.permission).toBeUndefined()
const broadBundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "broad",
})
expect(broadBundle.config.tools).toBeUndefined()
expect(broadBundle.config.permission).toEqual({
read: "allow",
write: "allow",
edit: "allow",
bash: "allow",
grep: "allow",
glob: "allow",
list: "allow",
webfetch: "allow",
skill: "allow",
patch: "allow",
task: "allow",
question: "allow",
todowrite: "allow",
todoread: "allow",
})
})
test("supports primary agent mode", async () => {
const plugin = await loadClaudePlugin(fixtureRoot)
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "primary",
inferTemperature: false,
permissions: "none",
})
const agentFile = bundle.agents.find((agent) => agent.name === "repo-research-analyst")
const parsed = parseFrontmatter(agentFile!.content)
expect(parsed.data.mode).toBe("primary")
})
test("excludes commands with disable-model-invocation from commandFiles", async () => {
const plugin = await loadClaudePlugin(fixtureRoot)
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "none",
})
// deploy-docs has disable-model-invocation: true, should be excluded
expect(bundle.commandFiles.find((f) => f.name === "deploy-docs")).toBeUndefined()
// Normal commands should still be present
expect(bundle.commandFiles.find((f) => f.name === "workflows:review")).toBeDefined()
})
test("rewrites .claude/ paths to .opencode/ in command bodies", () => {
const plugin: ClaudePlugin = {
root: "/tmp/plugin",
manifest: { name: "fixture", version: "1.0.0" },
agents: [],
commands: [
{
name: "review",
description: "Review command",
body: `Read \`compound-engineering.local.md\` in the project root.
If no settings file exists, auto-detect project type.
Run \`/compound-engineering-setup\` to create a settings file.`,
sourcePath: "/tmp/plugin/commands/review.md",
},
],
skills: [],
}
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "none",
})
const commandFile = bundle.commandFiles.find((f) => f.name === "review")
expect(commandFile).toBeDefined()
// Tool-agnostic path in project root — no rewriting needed
expect(commandFile!.content).toContain("compound-engineering.local.md")
})
test("rewrites .claude/ paths in agent bodies", () => {
const plugin: ClaudePlugin = {
root: "/tmp/plugin",
manifest: { name: "fixture", version: "1.0.0" },
agents: [
{
name: "test-agent",
description: "Test agent",
body: "Read `compound-engineering.local.md` for config.",
sourcePath: "/tmp/plugin/agents/test-agent.md",
},
],
commands: [],
skills: [],
}
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "none",
})
const agentFile = bundle.agents.find((a) => a.name === "test-agent")
expect(agentFile).toBeDefined()
// Tool-agnostic path in project root — no rewriting needed
expect(agentFile!.content).toContain("compound-engineering.local.md")
})
test("command .md files include description in frontmatter", () => {
const plugin: ClaudePlugin = {
root: "/tmp/plugin",
manifest: { name: "fixture", version: "1.0.0" },
agents: [],
commands: [
{
name: "test-cmd",
description: "Test description",
body: "Do the thing",
sourcePath: "/tmp/plugin/commands/test-cmd.md",
},
],
skills: [],
}
const bundle = convertClaudeToOpenCode(plugin, {
agentMode: "subagent",
inferTemperature: false,
permissions: "none",
})
const commandFile = bundle.commandFiles.find((f) => f.name === "test-cmd")
expect(commandFile).toBeDefined()
const parsed = parseFrontmatter(commandFile!.content)
expect(parsed.data.description).toBe("Test description")
expect(parsed.body).toContain("Do the thing")
})
})
describe("transformSkillContentForOpenCode", () => {
test("rewrites 3-segment FQ agent names to flat names", () => {
const input = "- `compound-engineering:document-review:coherence-reviewer`"
expect(transformSkillContentForOpenCode(input)).toBe("- `coherence-reviewer`")
})
test("rewrites multiple FQ agent refs in one block", () => {
const input = [
"- `compound-engineering:document-review:coherence-reviewer`",
"- `compound-engineering:document-review:feasibility-reviewer`",
"- `compound-engineering:review:security-sentinel`",
].join("\n")
const result = transformSkillContentForOpenCode(input)
expect(result).toContain("- `coherence-reviewer`")
expect(result).toContain("- `feasibility-reviewer`")
expect(result).toContain("- `security-sentinel`")
expect(result).not.toContain("compound-engineering:")
})
test("preserves 2-segment skill references", () => {
const input = 'load the `compound-engineering:document-review` skill'
// 2-segment refs are skill names, not agent names — left unchanged
expect(transformSkillContentForOpenCode(input)).toBe(input)
})
test("rewrites .claude/ paths to .opencode/", () => {
const input = "Read `.claude/config.json`"
expect(transformSkillContentForOpenCode(input)).toBe("Read `.opencode/config.json`")
})
test("rewrites ~/. claude/ paths to ~/.config/opencode/", () => {
const input = "Look in `~/.claude/plugins/`"
expect(transformSkillContentForOpenCode(input)).toBe("Look in `~/.config/opencode/plugins/`")
})
test("handles FQ names in JSON-like contexts", () => {
const input = ' subagent_type: "compound-engineering:review:security-sentinel",'
expect(transformSkillContentForOpenCode(input)).toBe(
' subagent_type: "security-sentinel",'
)
})
test("does not match URLs or non-agent colon patterns", () => {
const cases = [
"Visit https://example.com/path",
"Use http://localhost:8080/api",
"Set font-size: 12px; color: red;",
"Time is 10:30:45 UTC",
'key: "value"',
]
for (const input of cases) {
expect(transformSkillContentForOpenCode(input)).toBe(input)
}
})
test("rewrites FQ names from any plugin namespace", () => {
const input = "- `other-plugin:category:my-agent`"
expect(transformSkillContentForOpenCode(input)).toBe("- `my-agent`")
})
test("preserves bare agent names (no namespace)", () => {
const input = "Use `coherence-reviewer` for review."
expect(transformSkillContentForOpenCode(input)).toBe(input)
})
test("rewrites 2-segment category:ce-agent refs to flat names", () => {
const input = "Dispatch `review:ce-correctness-reviewer` for logic checks."
expect(transformSkillContentForOpenCode(input)).toBe(
"Dispatch `ce-correctness-reviewer` for logic checks.",
)
})
test("preserves 2-segment refs without ce- prefix", () => {
const input = "Spawn `compound-engineering:coherence-reviewer` as subagent."
// 2-segment names without ce- prefix could be skill refs — not rewritten
expect(transformSkillContentForOpenCode(input)).toBe(input)
})
test("does not partially rewrite 4-segment colon patterns", () => {
const input = "`a:b:c:d`"
// Without the lookahead, this would become `c:d` — a broken partial rewrite
expect(transformSkillContentForOpenCode(input)).toBe(input)
})
test("preserves 3-segment slash commands", () => {
const cases = [
"Run `/team:ops:deploy` to deploy.",
"Use /compound-engineering:review:check after changes.",
]
for (const input of cases) {
expect(transformSkillContentForOpenCode(input)).toBe(input)
}
})
})