Compare commits

...

15 Commits

Author SHA1 Message Date
xiaoju 36a3ca6a08 chore: bump cli@0.2.0, util@0.1.2
CI / check (push) Successful in 2m25s
2026-06-05 10:11:19 +00:00
xiaomo eb0b7b514f Merge pull request 'docs: update wf-stateless-design.md for new/resume $START semantics' (#103) from docs/101-stateless-design-update into main
CI / check (push) Successful in 2m9s
2026-06-05 09:49:23 +00:00
xiaoju a47871ec4e chore: remove unused moderator-reference and yaml-reference
CI / check (pull_request) Successful in 2m1s
These generate* functions were exported from util but never consumed
by any code. Dead exports are maintenance burden.

Refs #101
2026-06-05 09:44:50 +00:00
xiaoju 5851e5d162 docs: update wf-stateless-design.md to reflect new/resume semantics
CI / check (pull_request) Successful in 2m23s
Refs #101
2026-06-05 09:38:01 +00:00
xiaomo 61dfb40933 Merge pull request 'feat: replace $START _ status with new/resume semantics' (#102) from feat/101-start-new-resume into main
CI / check (push) Successful in 2m42s
2026-06-05 09:35:35 +00:00
xiaoju fbfd31a042 feat: replace $START _ status with new/resume semantics
CI / check (pull_request) Successful in 2m27s
BREAKING: All workflow YAML files must update $START._ to $START.new + $START.resume.
The resume edge prompt replaces the previously hardcoded resume message.

- evaluate.ts: remove START_ROLE/START_STATUS special case, use $status like all nodes
- thread.ts: resolveEvaluateArgs passes 'new', cmdThreadResume passes 'resume'
- validate.ts: reject '_' everywhere (no longer valid)
- validate-semantic.ts: require 'new' and 'resume' edges on $START
- All workflow YAMLs and test fixtures updated

Fixes #101
2026-06-05 09:30:09 +00:00
xiaomo d99a376b60 Merge pull request 'fix: simplify prompt subcommands, framework-agnostic bootstrap' (#100) from fix/99-prompt-cleanup into main
CI / check (push) Successful in 3m19s
2026-06-05 09:03:56 +00:00
xiaoju a536efee00 fix: simplify prompt subcommands, framework-agnostic bootstrap
CI / check (pull_request) Successful in 3m24s
- `uwf prompt usage` now outputs only the usage skill (was three combined)
- `uwf prompt bootstrap` replaces `setup` with framework-agnostic instructions
- Remove `usage-reference` and `setup` subcommands
- Remove `generateBootstrapReference` from util (moved to cli)

Fixes #99

小橘 🍊(NEKO Team)
2026-06-05 08:52:35 +00:00
xiaoju 9260d81084 chore: version bump for --version fix
CI / check (push) Successful in 3m2s
agent-hermes@0.1.2 agent-claude-code@0.1.1 agent-builtin@0.1.1
agent-mock@0.1.1 eval@0.1.3 util@0.1.1

小橘 🍊(NEKO Team)
2026-06-05 08:12:50 +00:00
xiaomo c8d884072a Merge pull request 'fix: acp-client reports agent-hermes own version in MCP clientInfo' (#98) from fix/acp-client-own-version into main
CI / check (push) Successful in 2m27s
2026-06-05 08:10:57 +00:00
xiaoju abeb465f46 fix: acp-client reports own package version, not util VERSION
CI / check (pull_request) Successful in 2m36s
Address review nit from PR #97: clientInfo.version should be
agent-hermes's own version for correct identification under
independent versioning.

小橘 🍊(NEKO Team)
2026-06-05 07:50:03 +00:00
xiaomo 28427a973f Merge pull request 'fix: add --version to adapter CLIs, read VERSION from package.json' (#97) from fix/adapter-version into main
CI / check (push) Successful in 3m3s
2026-06-05 07:36:15 +00:00
xiaoju 794f9db568 fix: add --version to adapter CLIs, read VERSION from package.json
CI / check (pull_request) Successful in 3m29s
- All uwf-* adapter CLIs now support --version / -V
- util VERSION constant reads from package.json at runtime
- agent-hermes ACP clientInfo uses dynamic VERSION

小橘 🍊(NEKO Team)
2026-06-05 07:29:54 +00:00
xiaoju cd585a26f1 Merge pull request 'fix: read eval CLI version from package.json' (#96) from fix/95-eval-version into main
CI / check (push) Successful in 3m28s
2026-06-05 06:46:32 +00:00
xiaoju 1cf8f350d0 fix: read eval CLI version from package.json
CI / check (pull_request) Successful in 3m30s
Fixes #95

小橘 🍊(NEKO Team)
2026-06-05 06:43:27 +00:00
54 changed files with 330 additions and 404 deletions
+9
View File
@@ -0,0 +1,9 @@
---
"@united-workforce/cli": minor
"@united-workforce/util": patch
---
feat: replace $START `_` status with `new`/`resume` semantics
BREAKING: All workflow YAML files must update `$START._` to `$START.new` + `$START.resume`.
The `resume` edge prompt replaces the previously hardcoded resume message in the CLI.
+2 -1
View File
@@ -264,7 +264,8 @@ roles:
graph: graph:
$START: $START:
_: { role: "bootstrap", prompt: "Set up the Docker container and verify uwf is runnable." } new: { role: "bootstrap", prompt: "Set up the Docker container and verify uwf is runnable." }
resume: { role: "bootstrap", prompt: "Review the previous run output and continue the walkthrough." }
bootstrap: bootstrap:
pass: { role: "config-and-registry", prompt: "Container {{{containerName}}} is ready. Validate config and workflow registration." } pass: { role: "config-and-registry", prompt: "Container {{{containerName}}} is ready. Validate config and workflow registration." }
fail: { role: "$END", prompt: "Bootstrap failed: {{{error}}}. No container was created." } fail: { role: "$END", prompt: "Bootstrap failed: {{{error}}}. No container was created." }
+2 -1
View File
@@ -227,7 +227,8 @@ roles:
required: [$status, error] required: [$status, error]
graph: graph:
$START: $START:
_: { role: "planner", prompt: "Analyze the issue and produce an implementation plan." } new: { role: "planner", prompt: "Analyze the issue and produce an implementation plan." }
resume: { role: "planner", prompt: "Review the previous run output and continue the work." }
planner: planner:
insufficient_info: { role: "$SUSPEND", prompt: "信息不足,需要补充:{{{reason}}}" } insufficient_info: { role: "$SUSPEND", prompt: "信息不足,需要补充:{{{reason}}}" }
ready: { role: "developer", prompt: "Implement the TDD test spec (CAS hash: {{{plan}}}) in repo {{{repoPath}}}. Repo remote: {{{repoRemote}}}." } ready: { role: "developer", prompt: "Implement the TDD test spec (CAS hash: {{{plan}}}) in repo {{{repoPath}}}. Repo remote: {{{repoRemote}}}." }
+4 -4
View File
@@ -200,7 +200,7 @@ payload:
- `roles` — 内联定义,每个 role 的 `meta` 是独立的 ocas_ref(指向 ocas 内置 JSON Schema 节点) - `roles` — 内联定义,每个 role 的 `meta` 是独立的 ocas_ref(指向 ocas 内置 JSON Schema 节点)
- `graph``Record<Role | "$START", Record<Status, Target>>`,每个 Target = `{ role, prompt }` - `graph``Record<Role | "$START", Record<Status, Target>>`,每个 Target = `{ role, prompt }`
- Status 来自上一个 role 输出的 `status` 字段,`$START``_` 作为初始 status - Status 来自上一个 role 输出的 `$status` 字段,`$START` 使`new`(首次启动)和 `resume`(恢复已完成的 thread)作为 status
- Prompt 模板使用 Mustache 渲染,变量来自 lastOutput - Prompt 模板使用 Mustache 渲染,变量来自 lastOutput
- 不含 agent binding — agent 配置在 `~/.uwf/config.yaml` 中管理 - 不含 agent binding — agent 配置在 `~/.uwf/config.yaml` 中管理
@@ -208,7 +208,7 @@ Moderator 的求值逻辑:
```typescript ```typescript
evaluate(graph, lastRole, lastOutput) { role, prompt } evaluate(graph, lastRole, lastOutput) { role, prompt }
// 1. status = lastRole === "$START" ? "_" : lastOutput.status // 1. status = lastOutput.$status (e.g. "new" for $START first run, "resume" for completed thread resume)
// 2. target = graph[lastRole][status] // 2. target = graph[lastRole][status]
// 3. prompt = mustache.render(target.prompt, lastOutput) // 3. prompt = mustache.render(target.prompt, lastOutput)
``` ```
@@ -422,8 +422,8 @@ type StepNodePayload = StepRecord & {
Moderator 使用 `evaluate(graph, lastRole, lastOutput)` 进行同步 status-based routing: Moderator 使用 `evaluate(graph, lastRole, lastOutput)` 进行同步 status-based routing:
```typescript ```typescript
// graph[lastRole][lastOutput.status] → Target { role, prompt } // graph[lastRole][lastOutput.$status] → Target { role, prompt }
// $START 角色使用 "_" 作为初始 status // $START 使用 "new"(首次启动)和 "resume"(恢复已完成 thread)作为 status
// prompt 通过 Mustache 模板渲染,变量来自 lastOutput // prompt 通过 Mustache 模板渲染,变量来自 lastOutput
``` ```
+2 -1
View File
@@ -35,6 +35,7 @@ roles:
required: [$status, thesis, keyPoints] required: [$status, thesis, keyPoints]
graph: graph:
$START: $START:
_: { role: "analyst", prompt: "Analyze the topic in the task and produce a structured summary with key points." } new: { role: "analyst", prompt: "Analyze the topic in the task and produce a structured summary with key points." }
resume: { role: "analyst", prompt: "Review the previous analysis output and continue with additional context." }
analyst: analyst:
done: { role: "$END", prompt: "Analysis complete. Finish the workflow." } done: { role: "$END", prompt: "Analysis complete. Finish the workflow." }
+2 -1
View File
@@ -53,7 +53,8 @@ roles:
required: [$status, argument] required: [$status, argument]
graph: graph:
$START: $START:
_: { role: "against", prompt: "Present your opening argument against the proposition." } new: { role: "against", prompt: "Present your opening argument against the proposition." }
resume: { role: "against", prompt: "Review the previous debate output and continue the argument against the proposition." }
against: against:
conceded: { role: "$END", prompt: "The against side conceded. Debate over." } conceded: { role: "$END", prompt: "The against side conceded. Debate over." }
continue: { role: "for", prompt: "Counter the opposing argument: {{{argument}}}" } continue: { role: "for", prompt: "Counter the opposing argument: {{{argument}}}" }
+2 -1
View File
@@ -25,6 +25,7 @@ roles:
required: [$status, summary] required: [$status, summary]
graph: graph:
$START: $START:
_: { role: "fixer", prompt: "Fix the code issue described in the task prompt." } new: { role: "fixer", prompt: "Fix the code issue described in the task prompt." }
resume: { role: "fixer", prompt: "Review the previous run output and continue fixing the code issue." }
fixer: fixer:
done: { role: "$END", prompt: "Fix complete." } done: { role: "$END", prompt: "Fix complete." }
+2 -1
View File
@@ -215,7 +215,8 @@ roles:
required: [$status, error] required: [$status, error]
graph: graph:
$START: $START:
_: { role: "planner", prompt: "Analyze the issue and produce an implementation plan." } new: { role: "planner", prompt: "Analyze the issue and produce an implementation plan." }
resume: { role: "planner", prompt: "Review the previous run output and continue the work." }
planner: planner:
insufficient_info: { role: "$SUSPEND", prompt: "信息不足,需要补充:{{{reason}}}" } insufficient_info: { role: "$SUSPEND", prompt: "信息不足,需要补充:{{{reason}}}" }
ready: { role: "developer", prompt: "Implement the TDD test spec (CAS hash: {{{plan}}}) in repo {{{repoPath}}}." } ready: { role: "developer", prompt: "Implement the TDD test spec (CAS hash: {{{plan}}}) in repo {{{repoPath}}}." }
+1 -1
View File
@@ -1,6 +1,6 @@
{ {
"name": "@united-workforce/agent-builtin", "name": "@united-workforce/agent-builtin",
"version": "0.1.0", "version": "0.1.1",
"files": [ "files": [
"src", "src",
"dist", "dist",
+7
View File
@@ -1,5 +1,12 @@
#!/usr/bin/env node #!/usr/bin/env node
// eslint-disable-next-line -- dynamic import for version
const pkg = await import("../package.json", { with: { type: "json" } });
if (process.argv.includes("--version") || process.argv.includes("-V")) {
process.stdout.write(`${pkg.default.version}\n`);
process.exit(0);
}
import { createBuiltinAgent } from "./agent.js"; import { createBuiltinAgent } from "./agent.js";
const main = createBuiltinAgent(); const main = createBuiltinAgent();
+1 -1
View File
@@ -1,6 +1,6 @@
{ {
"name": "@united-workforce/agent-claude-code", "name": "@united-workforce/agent-claude-code",
"version": "0.1.0", "version": "0.1.1",
"files": [ "files": [
"src", "src",
"dist", "dist",
+7
View File
@@ -1,5 +1,12 @@
#!/usr/bin/env node #!/usr/bin/env node
// eslint-disable-next-line -- dynamic import for version
const pkg = await import("../package.json", { with: { type: "json" } });
if (process.argv.includes("--version") || process.argv.includes("-V")) {
process.stdout.write(`${pkg.default.version}\n`);
process.exit(0);
}
import { createClaudeCodeAgent } from "./claude-code.js"; import { createClaudeCodeAgent } from "./claude-code.js";
const model = process.env.CLAUDE_MODEL ?? null; const model = process.env.CLAUDE_MODEL ?? null;
+1 -1
View File
@@ -1,6 +1,6 @@
{ {
"name": "@united-workforce/agent-hermes", "name": "@united-workforce/agent-hermes",
"version": "0.1.1", "version": "0.1.2",
"files": [ "files": [
"src", "src",
"dist", "dist",
+11 -1
View File
@@ -1,6 +1,16 @@
import type { ChildProcess } from "node:child_process"; import type { ChildProcess } from "node:child_process";
import { spawn } from "node:child_process"; import { spawn } from "node:child_process";
import { readFileSync } from "node:fs";
import { dirname, join } from "node:path";
import { createInterface } from "node:readline"; import { createInterface } from "node:readline";
import { fileURLToPath } from "node:url";
const __dirname = dirname(fileURLToPath(import.meta.url));
const OWN_VERSION = (
JSON.parse(readFileSync(join(__dirname, "..", "package.json"), "utf-8")) as {
version: string;
}
).version;
const HERMES_COMMAND = "hermes"; const HERMES_COMMAND = "hermes";
const PROTOCOL_VERSION = 1; const PROTOCOL_VERSION = 1;
@@ -299,7 +309,7 @@ export class HermesAcpClient {
private async initialize(): Promise<void> { private async initialize(): Promise<void> {
const initResponse = await this.sendRequest("initialize", { const initResponse = await this.sendRequest("initialize", {
protocolVersion: PROTOCOL_VERSION, protocolVersion: PROTOCOL_VERSION,
clientInfo: { name: "uwf", version: "0.1.0" }, clientInfo: { name: "uwf-hermes", version: OWN_VERSION },
capabilities: {}, capabilities: {},
}); });
+7
View File
@@ -1,5 +1,12 @@
#!/usr/bin/env node #!/usr/bin/env node
// eslint-disable-next-line -- dynamic import for version
const pkg = await import("../package.json", { with: { type: "json" } });
if (process.argv.includes("--version") || process.argv.includes("-V")) {
process.stdout.write(`${pkg.default.version}\n`);
process.exit(0);
}
import { createHermesAgent } from "./hermes.js"; import { createHermesAgent } from "./hermes.js";
import { isResumeDisabled } from "./session-cache.js"; import { isResumeDisabled } from "./session-cache.js";
+1 -1
View File
@@ -1,6 +1,6 @@
{ {
"name": "@united-workforce/agent-mock", "name": "@united-workforce/agent-mock",
"version": "0.1.0", "version": "0.1.1",
"files": [ "files": [
"src", "src",
"dist", "dist",
+7
View File
@@ -1,5 +1,12 @@
#!/usr/bin/env node #!/usr/bin/env node
// eslint-disable-next-line -- dynamic import for version
const pkg = await import("../package.json", { with: { type: "json" } });
if (process.argv.includes("--version") || process.argv.includes("-V")) {
process.stdout.write(`${pkg.default.version}\n`);
process.exit(0);
}
import { createMockAgent } from "./mock-agent.js"; import { createMockAgent } from "./mock-agent.js";
const USAGE = "usage: uwf-mock --mock-data <path> --thread <id> --role <role> --prompt <text>"; const USAGE = "usage: uwf-mock --mock-data <path> --thread <id> --role <role> --prompt <text>";
+1 -1
View File
@@ -1,6 +1,6 @@
{ {
"name": "@united-workforce/cli", "name": "@united-workforce/cli",
"version": "0.1.1", "version": "0.2.0",
"files": [ "files": [
"src", "src",
"dist", "dist",
@@ -58,7 +58,10 @@ describe("C1: adapter JSON round-trip integration", () => {
}, },
}, },
graph: { graph: {
$START: { _: { role: "worker", prompt: "Do the work", location: null } }, $START: {
new: { role: "worker", prompt: "Do the work", location: null },
resume: { role: "worker", prompt: "Resume the work", location: null },
},
worker: { done: { role: "$END", prompt: "completed", location: null } }, worker: { done: { role: "$END", prompt: "completed", location: null } },
}, },
}); });
@@ -45,10 +45,14 @@ roles:
$status: { type: string, enum: ["done"] } $status: { type: string, enum: ["done"] }
graph: graph:
$START: $START:
_: new:
role: roleA role: roleA
prompt: "Do A" prompt: "Do A"
location: null location: null
resume:
role: roleA
prompt: "Resume A"
location: null
roleA: roleA:
ready: ready:
role: roleB role: roleB
@@ -107,10 +111,14 @@ roles:
$status: { type: string, enum: ["done"] } $status: { type: string, enum: ["done"] }
graph: graph:
$START: $START:
_: new:
role: roleA role: roleA
prompt: "Do A" prompt: "Do A"
location: null location: null
resume:
role: roleA
prompt: "Resume A"
location: null
roleA: roleA:
pass: pass:
role: roleB role: roleB
@@ -150,10 +158,14 @@ roles:
$status: { type: string, enum: ["done"] } $status: { type: string, enum: ["done"] }
graph: graph:
$START: $START:
_: new:
role: worker role: worker
prompt: "Work" prompt: "Work"
location: null location: null
resume:
role: worker
prompt: "Resume work"
location: null
worker: worker:
done: done:
role: $END role: $END
@@ -36,7 +36,8 @@ roles:
required: [$status] required: [$status]
graph: graph:
$START: $START:
_: { role: analyst, prompt: 'Analyze the task' } new: { role: analyst, prompt: 'Analyze the task' }
resume: { role: analyst, prompt: 'Review the previous run output and continue the work.' }
analyst: analyst:
analyzed: { role: developer, prompt: 'Implement the change' } analyzed: { role: developer, prompt: 'Implement the change' }
developer: developer:
@@ -25,7 +25,8 @@ roles:
required: [$status] required: [$status]
graph: graph:
$START: $START:
_: { role: planner, prompt: 'Plan the task' } new: { role: planner, prompt: 'Plan the task' }
resume: { role: planner, prompt: 'Review the previous run output and continue the work.' }
planner: planner:
ready: { role: worker, prompt: 'Do the work' } ready: { role: worker, prompt: 'Do the work' }
worker: worker:
@@ -28,7 +28,8 @@ roles:
required: [$status] required: [$status]
graph: graph:
$START: $START:
_: { role: developer, prompt: 'Implement the change' } new: { role: developer, prompt: 'Implement the change' }
resume: { role: developer, prompt: 'Review the previous run output and continue the work.' }
developer: developer:
review_needed: { role: reviewer, prompt: 'Review the change' } review_needed: { role: reviewer, prompt: 'Review the change' }
reviewer: reviewer:
@@ -27,7 +27,8 @@ roles:
required: [$status] required: [$status]
graph: graph:
$START: $START:
_: { role: planner, prompt: 'Plan the task' } new: { role: planner, prompt: 'Plan the task' }
resume: { role: planner, prompt: 'Review the previous run output and continue the work.' }
planner: planner:
ready: { role: worker, prompt: 'Work on branch {{{branch}}} in {{{repoPath}}}' } ready: { role: worker, prompt: 'Work on branch {{{branch}}} in {{{repoPath}}}' }
worker: worker:
@@ -18,7 +18,8 @@ roles:
required: [$status] required: [$status]
graph: graph:
$START: $START:
_: { role: planner, prompt: 'Analyze the task' } new: { role: planner, prompt: 'Analyze the task' }
resume: { role: planner, prompt: 'Review the previous run output and continue the work.' }
planner: planner:
insufficient_info: { role: '$SUSPEND', prompt: 'Need more info: {{{reason}}}' } insufficient_info: { role: '$SUSPEND', prompt: 'Need more info: {{{reason}}}' }
ready: { role: '$END', prompt: 'Done' } ready: { role: '$END', prompt: 'Done' }
@@ -5,7 +5,12 @@ import { evaluate } from "../moderator/evaluate.js";
const solveIssueGraph: WorkflowPayload["graph"] = { const solveIssueGraph: WorkflowPayload["graph"] = {
$START: { $START: {
_: { role: "planner", prompt: "Start planning from the issue in the task.", location: null }, new: { role: "planner", prompt: "Start planning from the issue in the task.", location: null },
resume: {
role: "planner",
prompt: "Review the previous run output and continue the work.",
location: null,
},
}, },
planner: { planner: {
planned: { role: "developer", prompt: "Implement the plan: {{plan}}", location: null }, planned: { role: "developer", prompt: "Implement the plan: {{plan}}", location: null },
@@ -20,8 +25,8 @@ const solveIssueGraph: WorkflowPayload["graph"] = {
}; };
describe("evaluate", () => { describe("evaluate", () => {
test("$START → first role (unit status _)", () => { test("$START → first role (status new)", () => {
const result = evaluate(solveIssueGraph, "$START", { $status: "_" }); const result = evaluate(solveIssueGraph, "$START", { $status: "new" });
expect(result).toEqual({ expect(result).toEqual({
ok: true, ok: true,
value: { value: {
@@ -32,6 +37,18 @@ describe("evaluate", () => {
}); });
}); });
test("$START → first role (status resume)", () => {
const result = evaluate(solveIssueGraph, "$START", { $status: "resume" });
expect(result).toEqual({
ok: true,
value: {
role: "planner",
prompt: "Review the previous run output and continue the work.",
location: null,
},
});
});
test("status-based routing (reviewer rejected → developer)", () => { test("status-based routing (reviewer rejected → developer)", () => {
const result = evaluate(solveIssueGraph, "reviewer", { const result = evaluate(solveIssueGraph, "reviewer", {
$status: "rejected", $status: "rejected",
@@ -95,7 +112,7 @@ describe("evaluate", () => {
}); });
test("missing role in graph → error", () => { test("missing role in graph → error", () => {
const result = evaluate(solveIssueGraph, "unknown-role", { $status: "_" }); const result = evaluate(solveIssueGraph, "unknown-role", { $status: "new" });
expect(result.ok).toBe(false); expect(result.ok).toBe(false);
if (!result.ok) { if (!result.ok) {
expect(result.error.message).toBe('no transitions defined for role "unknown-role"'); expect(result.error.message).toBe('no transitions defined for role "unknown-role"');
+19 -46
View File
@@ -9,31 +9,25 @@ import {
cmdPromptAdapterDeveloping, cmdPromptAdapterDeveloping,
cmdPromptBootstrap, cmdPromptBootstrap,
cmdPromptList, cmdPromptList,
cmdPromptSetup,
cmdPromptUsage, cmdPromptUsage,
cmdPromptUsageReference,
cmdPromptWorkflowAuthoring, cmdPromptWorkflowAuthoring,
} from "../commands/prompt.js"; } from "../commands/prompt.js";
describe("prompt commands", () => { describe("prompt commands", () => {
test("prompt list returns new prompt names", () => { test("prompt list returns prompt names (no bootstrap)", () => {
const result = cmdPromptList(); const result = cmdPromptList();
expect(result).toBeInstanceOf(Array); expect(result).toBeInstanceOf(Array);
expect(result).toContain("usage"); expect(result).toContain("usage");
expect(result).toContain("workflow-authoring"); expect(result).toContain("workflow-authoring");
expect(result).toContain("adapter-developing"); expect(result).toContain("adapter-developing");
expect(result).toContain("bootstrap"); expect(result).not.toContain("bootstrap");
expect(result).not.toContain("user");
expect(result).not.toContain("author");
expect(result).not.toContain("developer");
expect(result).not.toContain("adapter");
for (const name of result) { for (const name of result) {
expect(name).toMatch(/^\S+$/); expect(name).toMatch(/^\S+$/);
} }
}); });
test("prompt usage-reference returns non-empty markdown string with frontmatter", () => { test("prompt usage returns only the usage reference with frontmatter", () => {
const result = cmdPromptUsageReference(); const result = cmdPromptUsage();
expect(typeof result).toBe("string"); expect(typeof result).toBe("string");
expect(result).toContain("uwf"); expect(result).toContain("uwf");
expect(result).toContain("thread"); expect(result).toContain("thread");
@@ -42,6 +36,9 @@ describe("prompt commands", () => {
expect(result).toContain("---"); expect(result).toContain("---");
expect(result).toContain("name:"); expect(result).toContain("name:");
expect(result).toContain("version:"); expect(result).toContain("version:");
// Should NOT contain other references
expect(result).not.toContain("Workflow Authoring Reference");
expect(result).not.toContain("Adapter Developing Reference");
expect(result.length).toBeGreaterThan(500); expect(result.length).toBeGreaterThan(500);
}); });
@@ -71,44 +68,19 @@ describe("prompt commands", () => {
expect(result.length).toBeGreaterThan(500); expect(result.length).toBeGreaterThan(500);
}); });
test("prompt bootstrap returns non-empty skill with frontmatter", () => { test("prompt bootstrap returns framework-agnostic setup instructions", () => {
const result = cmdPromptBootstrap(); const result = cmdPromptBootstrap();
expect(typeof result).toBe("string"); expect(typeof result).toBe("string");
expect(result).toContain("uwf");
expect(result).toContain("---");
expect(result.length).toBeGreaterThan(100);
});
test("prompt usage combines remaining references (no developer)", () => {
const result = cmdPromptUsage();
expect(typeof result).toBe("string");
expect(result).toContain("Usage Reference");
expect(result).toContain("Workflow Authoring Reference");
expect(result).toContain("Adapter Developing Reference");
expect(result).not.toContain("Developer Reference");
expect(result).toContain("---");
expect(result.length).toBeGreaterThan(2000);
});
test("prompt setup returns simplified setup instructions", () => {
const result = cmdPromptSetup();
expect(typeof result).toBe("string");
expect(result).toContain("uwf Skill Setup");
expect(result).toContain("uwf prompt bootstrap");
expect(result).toContain("SKILL.md");
expect(result).toContain("version");
expect(result).not.toMatch(/\bbun (install|run|test|changeset|version|release)\b/);
});
test("prompt setup references new subcommand names", () => {
const result = cmdPromptSetup();
expect(result).toContain("uwf prompt usage"); expect(result).toContain("uwf prompt usage");
expect(result).toContain("uwf prompt workflow-authoring"); expect(result).toContain("uwf prompt workflow-authoring");
expect(result).toContain("uwf prompt adapter-developing"); expect(result).toContain("uwf prompt adapter-developing");
expect(result).not.toContain("uwf prompt user"); expect(result).toContain("uwf-usage");
expect(result).not.toContain("uwf prompt author"); expect(result).toContain("uwf-workflow-authoring");
expect(result).not.toContain("uwf prompt developer"); expect(result).toContain("uwf-adapter-developing");
expect(result).not.toMatch(/uwf prompt adapter\b(?!-developing)/); // Should NOT contain Hermes-specific paths
expect(result).not.toContain("~/.hermes/skills/");
expect(result).not.toContain("> ~/.hermes/");
expect(result.length).toBeGreaterThan(100);
}); });
test("prompt help subcommand is suppressed", { timeout: 30_000 }, () => { test("prompt help subcommand is suppressed", { timeout: 30_000 }, () => {
@@ -119,11 +91,12 @@ describe("prompt commands", () => {
}); });
expect(output).not.toMatch(/help\s+\[command\]/i); expect(output).not.toMatch(/help\s+\[command\]/i);
expect(output).toContain("usage"); expect(output).toContain("usage");
expect(output).toContain("setup"); expect(output).toContain("bootstrap");
expect(output).toContain("workflow-authoring"); expect(output).toContain("workflow-authoring");
expect(output).toContain("adapter-developing"); expect(output).toContain("adapter-developing");
expect(output).toContain("bootstrap");
expect(output).toContain("list"); expect(output).toContain("list");
expect(output).not.toContain("developer"); // Removed subcommands should not appear as command names
expect(output).not.toMatch(/^\s+setup\s/m);
expect(output).not.toContain("usage-reference");
}); });
}); });
@@ -253,7 +253,10 @@ describe("thread read timing", () => {
}, },
}, },
graph: { graph: {
$START: { _: { role: "worker", prompt: "go", location: null } }, $START: {
new: { role: "worker", prompt: "go", location: null },
resume: { role: "worker", prompt: "resume", location: null },
},
worker: { done: { role: "$END", prompt: "", location: null } }, worker: { done: { role: "$END", prompt: "", location: null } },
}, },
}); });
@@ -319,7 +322,10 @@ describe("thread read timing", () => {
}, },
}, },
graph: { graph: {
$START: { _: { role: "worker", prompt: "go", location: null } }, $START: {
new: { role: "worker", prompt: "go", location: null },
resume: { role: "worker", prompt: "resume", location: null },
},
worker: { done: { role: "$END", prompt: "", location: null } }, worker: { done: { role: "$END", prompt: "", location: null } },
}, },
}); });
@@ -57,10 +57,14 @@ roles:
$status: { type: string, enum: ["ready"] } $status: { type: string, enum: ["ready"] }
graph: graph:
$START: $START:
_: new:
role: planner role: planner
prompt: "Plan the work" prompt: "Plan the work"
location: null location: null
resume:
role: planner
prompt: "Resume the work"
location: null
planner: planner:
ready: ready:
role: $END role: $END
@@ -113,10 +117,14 @@ roles:
$status: { type: string, enum: ["ready"] } $status: { type: string, enum: ["ready"] }
graph: graph:
$START: $START:
_: new:
role: planner role: planner
prompt: "Plan" prompt: "Plan"
location: null location: null
resume:
role: planner
prompt: "Resume"
location: null
planner: planner:
ready: ready:
role: $END role: $END
@@ -156,10 +164,14 @@ roles:
$status: { type: string, enum: ["ready"] } $status: { type: string, enum: ["ready"] }
graph: graph:
$START: $START:
_: new:
role: planner role: planner
prompt: "Plan" prompt: "Plan"
location: null location: null
resume:
role: planner
prompt: "Resume"
location: null
planner: planner:
ready: ready:
role: $END role: $END
@@ -70,7 +70,10 @@ async function setupSuspendedThread(mode: MockAgentMode): Promise<{
}, },
}, },
graph: { graph: {
$START: { _: { role: "worker", prompt: "Start work", location: null } }, $START: {
new: { role: "worker", prompt: "Start work", location: null },
resume: { role: "worker", prompt: "Resume the work", location: null },
},
worker: { worker: {
needs_input: { needs_input: {
role: "$SUSPEND", role: "$SUSPEND",
@@ -233,7 +236,10 @@ describe("uwf thread resume", () => {
}, },
}, },
graph: { graph: {
$START: { _: { role: "worker", prompt: "Start", location: null } }, $START: {
new: { role: "worker", prompt: "Start", location: null },
resume: { role: "worker", prompt: "Resume", location: null },
},
worker: { done: { role: "$END", prompt: "Done", location: null } }, worker: { done: { role: "$END", prompt: "Done", location: null } },
}, },
}); });
@@ -479,7 +485,10 @@ describe("uwf thread resume - completed threads", () => {
}, },
}, },
graph: { graph: {
$START: { _: { role: "worker", prompt: "Start work", location: null } }, $START: {
new: { role: "worker", prompt: "Start work", location: null },
resume: { role: "worker", prompt: "Resume the work", location: null },
},
worker: { done: { role: "reviewer", prompt: "Review the work", location: null } }, worker: { done: { role: "reviewer", prompt: "Review the work", location: null } },
reviewer: { done: { role: "$END", prompt: "Done", location: null } }, reviewer: { done: { role: "$END", prompt: "Done", location: null } },
}, },
@@ -610,7 +619,7 @@ echo '${adapterJson}'
expect(cliOutput.done).toBe(false); expect(cliOutput.done).toBe(false);
const capturedPrompt = await readFile(promptCapturePath, "utf8"); const capturedPrompt = await readFile(promptCapturePath, "utf8");
expect(capturedPrompt).toContain("Previous run completed"); expect(capturedPrompt).toContain("Resume the work");
expect(capturedPrompt).toContain("Additional context"); expect(capturedPrompt).toContain("Additional context");
const storeModule = await import("../store.js"); const storeModule = await import("../store.js");
@@ -640,7 +649,10 @@ echo '${adapterJson}'
}, },
}, },
graph: { graph: {
$START: { _: { role: "worker", prompt: "Start", location: null } }, $START: {
new: { role: "worker", prompt: "Start", location: null },
resume: { role: "worker", prompt: "Resume", location: null },
},
worker: { done: { role: "$END", prompt: "Done", location: null } }, worker: { done: { role: "$END", prompt: "Done", location: null } },
}, },
}); });
@@ -688,7 +700,10 @@ echo '${adapterJson}'
}, },
}, },
graph: { graph: {
$START: { _: { role: "worker", prompt: "Start", location: null } }, $START: {
new: { role: "worker", prompt: "Start", location: null },
resume: { role: "worker", prompt: "Resume", location: null },
},
worker: { done: { role: "$END", prompt: "Done", location: null } }, worker: { done: { role: "$END", prompt: "Done", location: null } },
}, },
}); });
@@ -34,10 +34,14 @@ roles:
$status: { type: string, enum: ["ready"] } $status: { type: string, enum: ["ready"] }
graph: graph:
$START: $START:
_: new:
role: planner role: planner
prompt: "Plan the work" prompt: "Plan the work"
location: null location: null
resume:
role: planner
prompt: "Resume the work"
location: null
planner: planner:
ready: ready:
role: $END role: $END
@@ -66,10 +70,14 @@ roles:
question: { type: string } question: { type: string }
graph: graph:
$START: $START:
_: new:
role: worker role: worker
prompt: "Start work" prompt: "Start work"
location: null location: null
resume:
role: worker
prompt: "Resume work"
location: null
worker: worker:
needs_input: needs_input:
role: $SUSPEND role: $SUSPEND
@@ -57,10 +57,14 @@ roles:
$status: { type: string, enum: ["ready"] } $status: { type: string, enum: ["ready"] }
graph: graph:
$START: $START:
_: new:
role: planner role: planner
prompt: "Plan the work" prompt: "Plan the work"
location: null location: null
resume:
role: planner
prompt: "Resume the work"
location: null
planner: planner:
ready: ready:
role: $END role: $END
@@ -58,7 +58,10 @@ describe("suspend step CAS chain and threads.yaml metadata", () => {
}, },
}, },
graph: { graph: {
$START: { _: { role: "worker", prompt: "Start work", location: null } }, $START: {
new: { role: "worker", prompt: "Start work", location: null },
resume: { role: "worker", prompt: "Resume work", location: null },
},
worker: { worker: {
needs_input: { needs_input: {
role: "$SUSPEND", role: "$SUSPEND",
@@ -55,7 +55,10 @@ describe("suspended thread display", () => {
}, },
}, },
graph: { graph: {
$START: { _: { role: "worker", prompt: "Start work", location: null } }, $START: {
new: { role: "worker", prompt: "Start work", location: null },
resume: { role: "worker", prompt: "Resume work", location: null },
},
worker: { worker: {
needs_input: { needs_input: {
role: "$SUSPEND", role: "$SUSPEND",
@@ -162,7 +165,10 @@ describe("suspended thread display", () => {
}, },
}, },
graph: { graph: {
$START: { _: { role: "worker", prompt: "Start work", location: null } }, $START: {
new: { role: "worker", prompt: "Start work", location: null },
resume: { role: "worker", prompt: "Resume work", location: null },
},
worker: { worker: {
needs_input: { needs_input: {
role: "$SUSPEND", role: "$SUSPEND",
@@ -248,7 +254,10 @@ describe("suspended thread display", () => {
}, },
}, },
graph: { graph: {
$START: { _: { role: "worker", prompt: "Start work", location: null } }, $START: {
new: { role: "worker", prompt: "Start work", location: null },
resume: { role: "worker", prompt: "Resume work", location: null },
},
}, },
}); });
@@ -51,7 +51,10 @@ function makeWorkflow(overrides?: Partial<WorkflowPayload>): WorkflowPayload {
}, },
}, },
graph: { graph: {
$START: { _: { role: "writer", prompt: "Begin writing", location: null } }, $START: {
new: { role: "writer", prompt: "Begin writing", location: null },
resume: { role: "writer", prompt: "Review previous output and continue", location: null },
},
writer: { done: { role: "reviewer", prompt: "Review this: {{{plan}}}", location: null } }, writer: { done: { role: "reviewer", prompt: "Review this: {{{plan}}}", location: null } },
reviewer: { reviewer: {
approved: { role: "$END", prompt: "Done: {{{summary}}}", location: null }, approved: { role: "$END", prompt: "Done: {{{summary}}}", location: null },
@@ -135,27 +138,38 @@ describe("Suite 2: Graph Structure", () => {
expect(errors.some((e) => e.includes("$START must be defined in graph"))).toBe(true); expect(errors.some((e) => e.includes("$START must be defined in graph"))).toBe(true);
}); });
test("2.2 $START has multiple status keys", () => { test("2.2 $START missing resume edge", () => {
const wf = makeWorkflow(); const wf = makeWorkflow();
wf.graph.$START = { wf.graph.$START = {
_: { role: "writer", prompt: "Begin", location: null }, new: { role: "writer", prompt: "Begin", location: null },
other: { role: "reviewer", prompt: "Also", location: null },
}; };
const errors = validateWorkflow(wf); const errors = validateWorkflow(wf);
expect( expect(
errors.some((e) => e.includes('$START must have exactly one edge with status "_"')), errors.some((e) => e.includes('$START must have edges with statuses "new" and "resume"')),
).toBe(true); ).toBe(true);
}); });
test("2.3 $START edge uses non-_ status", () => { test("2.3 $START missing new edge", () => {
const wf = makeWorkflow(); const wf = makeWorkflow();
wf.graph.$START = { ready: { role: "writer", prompt: "Begin", location: null } }; wf.graph.$START = {
resume: { role: "writer", prompt: "Resume", location: null },
};
const errors = validateWorkflow(wf); const errors = validateWorkflow(wf);
expect( expect(
errors.some((e) => e.includes('$START must have exactly one edge with status "_"')), errors.some((e) => e.includes('$START must have edges with statuses "new" and "resume"')),
).toBe(true); ).toBe(true);
}); });
test("2.3b $START with new and resume passes", () => {
const wf = makeWorkflow();
wf.graph.$START = {
new: { role: "writer", prompt: "Begin", location: null },
resume: { role: "writer", prompt: "Resume", location: null },
};
const errors = validateWorkflow(wf);
expect(errors.some((e) => e.includes("$START must have edges"))).toBe(false);
});
test("2.4 $END has outgoing edges", () => { test("2.4 $END has outgoing edges", () => {
const wf = makeWorkflow(); const wf = makeWorkflow();
wf.graph.$END = { _: { role: "writer", prompt: "Loop", location: null } }; wf.graph.$END = { _: { role: "writer", prompt: "Loop", location: null } };
@@ -193,15 +207,18 @@ describe("Suite 2: Graph Structure", () => {
}); });
describe("Suite 3: Status-Edge Consistency", () => { describe("Suite 3: Status-Edge Consistency", () => {
test("3.1 user role using _ graph key is rejected", () => { test("3.1 user role using _ graph key is treated as an unknown status", () => {
// "_" is no longer special-cased — it's just a status key that does not
// match the role's $status enum, so it surfaces as extra/missing keys.
const wf = makeWorkflow(); const wf = makeWorkflow();
wf.graph.writer = { _: { role: "reviewer", prompt: "Review", location: null } }; wf.graph.writer = { _: { role: "reviewer", prompt: "Review", location: null } };
const errors = validateWorkflow(wf); const errors = validateWorkflow(wf);
expect( expect(errors.some((e) => e.includes('role "writer" graph has extra status keys: _'))).toBe(
errors.some((e) => true,
e.includes('role "writer" must use explicit $status keys in graph, not "_"'), );
), expect(errors.some((e) => e.includes('role "writer" graph is missing status keys: done'))).toBe(
).toBe(true); true,
);
}); });
test("3.2 user role graph key not matching $status enum", () => { test("3.2 user role graph key not matching $status enum", () => {
@@ -240,13 +257,16 @@ describe("Suite 3: Status-Edge Consistency", () => {
).toBe(true); ).toBe(true);
}); });
test("3.5 multi-exit role with _ key", () => { test("3.5 multi-exit role with _ key is treated as an unknown status", () => {
const wf = makeWorkflow(); const wf = makeWorkflow();
wf.graph.reviewer = { _: { role: "$END", prompt: "Done", location: null } }; wf.graph.reviewer = { _: { role: "$END", prompt: "Done", location: null } };
const errors = validateWorkflow(wf); const errors = validateWorkflow(wf);
expect(errors.some((e) => e.includes('role "reviewer" graph has extra status keys: _'))).toBe(
true,
);
expect( expect(
errors.some((e) => errors.some((e) =>
e.includes('role "reviewer" must use explicit $status keys in graph, not "_"'), e.includes('role "reviewer" graph is missing status keys: approved, rejected'),
), ),
).toBe(true); ).toBe(true);
}); });
@@ -38,7 +38,10 @@ function makeMinimalPayload(name: string, description: string): WorkflowPayload
}, },
}, },
graph: { graph: {
$START: { _: { role: "worker", prompt: "start working", location: null } }, $START: {
new: { role: "worker", prompt: "start working", location: null },
resume: { role: "worker", prompt: "resume working", location: null },
},
worker: { done: { role: "$END", prompt: "done", location: null } }, worker: { done: { role: "$END", prompt: "done", location: null } },
}, },
}; };
+4 -20
View File
@@ -8,9 +8,7 @@ import {
cmdPromptAdapterDeveloping, cmdPromptAdapterDeveloping,
cmdPromptBootstrap, cmdPromptBootstrap,
cmdPromptList, cmdPromptList,
cmdPromptSetup,
cmdPromptUsage, cmdPromptUsage,
cmdPromptUsageReference,
cmdPromptWorkflowAuthoring, cmdPromptWorkflowAuthoring,
} from "./commands/prompt.js"; } from "./commands/prompt.js";
import { cmdSetup, cmdSetupInteractive } from "./commands/setup.js"; import { cmdSetup, cmdSetupInteractive } from "./commands/setup.js";
@@ -509,23 +507,16 @@ prompt.addHelpCommand(false);
prompt prompt
.command("usage") .command("usage")
.description("Print the complete skill content (all references combined)") .description("Print the usage reference (CLI guide + typical workflows)")
.action(() => { .action(() => {
console.log(cmdPromptUsage()); console.log(cmdPromptUsage());
}); });
prompt prompt
.command("setup") .command("bootstrap")
.description("Print setup instructions for installing the uwf skill") .description("Print setup instructions for installing uwf skills")
.action(() => { .action(() => {
console.log(cmdPromptSetup()); console.log(cmdPromptBootstrap());
});
prompt
.command("usage-reference")
.description("Print the usage reference (CLI guide + typical workflows)")
.action(() => {
console.log(cmdPromptUsageReference());
}); });
prompt prompt
@@ -542,13 +533,6 @@ prompt
console.log(cmdPromptAdapterDeveloping()); console.log(cmdPromptAdapterDeveloping());
}); });
prompt
.command("bootstrap")
.description("Print the bootstrap skill YAML for Hermes agents")
.action(() => {
console.log(cmdPromptBootstrap());
});
prompt prompt
.command("list") .command("list")
.description("List all available prompt names") .description("List all available prompt names")
+22 -41
View File
@@ -1,14 +1,13 @@
import { import {
generateAdapterDevelopingReference, generateAdapterDevelopingReference,
generateBootstrapReference,
generateUsageReference, generateUsageReference,
generateWorkflowAuthoringReference, generateWorkflowAuthoringReference,
VERSION,
} from "@united-workforce/util"; } from "@united-workforce/util";
export { export {
generateAdapterDevelopingReference as cmdPromptAdapterDeveloping, generateAdapterDevelopingReference as cmdPromptAdapterDeveloping,
generateBootstrapReference as cmdPromptBootstrap, generateUsageReference as cmdPromptUsage,
generateUsageReference as cmdPromptUsageReference,
generateWorkflowAuthoringReference as cmdPromptWorkflowAuthoring, generateWorkflowAuthoringReference as cmdPromptWorkflowAuthoring,
}; };
@@ -16,24 +15,16 @@ const PROMPT_ENTRIES: ReadonlyArray<{ name: string; generate: () => string }> =
{ name: "usage", generate: generateUsageReference }, { name: "usage", generate: generateUsageReference },
{ name: "workflow-authoring", generate: generateWorkflowAuthoringReference }, { name: "workflow-authoring", generate: generateWorkflowAuthoringReference },
{ name: "adapter-developing", generate: generateAdapterDevelopingReference }, { name: "adapter-developing", generate: generateAdapterDevelopingReference },
{ name: "bootstrap", generate: generateBootstrapReference },
]; ];
export function cmdPromptList(): ReadonlyArray<string> { export function cmdPromptList(): ReadonlyArray<string> {
return PROMPT_ENTRIES.map((e) => e.name); return PROMPT_ENTRIES.map((e) => e.name);
} }
export function cmdPromptUsage(): string { export function cmdPromptBootstrap(): string {
return PROMPT_ENTRIES.filter((e) => e.name !== "bootstrap") return `# uwf Skill Bootstrap
.map((e) => e.generate())
.join("\n\n---\n\n");
}
export function cmdPromptSetup(): string { Install or update the uwf (United Workforce) skills for your agent.
return `# uwf Skill Setup
You are being asked to install or update the uwf (United Workforce) skill
so that you know how to use the \`uwf\` CLI for workflow orchestration.
## Steps ## Steps
@@ -42,40 +33,30 @@ so that you know how to use the \`uwf\` CLI for workflow orchestration.
uwf --help uwf --help
\`\`\` \`\`\`
If not installed: \`npm install -g @united-workforce/cli\` If not installed: \`npm install -g @united-workforce/cli\`
Current version: ${VERSION}
2. **Clean up old versions of the skill** (if any exist): 2. **Install the skills** by running each command below and creating a skill from its output:
- Look for any existing uwf/workflow skill files in your skill directories - Run \`uwf prompt usage\` → create a skill named \`uwf-usage\`
(e.g. \`~/.hermes/skills/uwf/SKILL.md\` or similar paths). - Run \`uwf prompt workflow-authoring\` → create a skill named \`uwf-workflow-authoring\`
- If a skill exists, read its \`version\` from the YAML frontmatter. - Run \`uwf prompt adapter-developing\` → create a skill named \`uwf-adapter-developing\`
- If the version matches the current CLI version (\`uwf --version\`), you're done — no update needed.
- If the version is older, **delete the old skill file** before proceeding.
3. **Install the bootstrap skill:** Each command outputs a complete SKILL.md with YAML frontmatter — use your agent framework's skill creation API to save them.
\`\`\`bash
uwf prompt bootstrap > ~/.hermes/skills/uwf/SKILL.md
\`\`\`
The bootstrap prompt already includes complete YAML frontmatter — no editing needed.
4. **Verify** the skill is loadable by your agent framework. 3. **Verify** the skills are loadable by your agent framework.
## Individual prompts ## Updating
Each prompt outputs a complete SKILL.md with frontmatter — pipe directly to a file: When \`uwf\` is upgraded, re-run \`uwf prompt bootstrap\` and follow the steps again.
The skill content is bundled with the CLI — always use \`uwf prompt <name>\` to get
content matching your installed version.
## Available prompts
\`\`\`bash \`\`\`bash
uwf prompt list # list available prompt names uwf prompt list # list available prompt names
uwf prompt usage > ~/.hermes/skills/uwf-usage/SKILL.md # CLI usage guide uwf prompt usage # CLI usage guide
uwf prompt workflow-authoring > ~/.hermes/skills/uwf-workflow-authoring/SKILL.md uwf prompt workflow-authoring # workflow YAML design guide
uwf prompt adapter-developing > ~/.hermes/skills/uwf-adapter-developing/SKILL.md uwf prompt adapter-developing # building agent adapters
uwf prompt bootstrap > ~/.hermes/skills/uwf/SKILL.md # bootstrap skill
\`\`\` \`\`\`
## Notes
- The skill content is bundled with the CLI and versioned with it — always use
\`uwf prompt usage\` to get the content matching your installed version.
- Do NOT hand-edit the skill body. If the CLI is updated, re-run \`uwf prompt setup\`
and follow the steps again.
- When upgrading, always delete the old skill first to avoid stale instructions.
`; `;
} }
+3 -8
View File
@@ -911,7 +911,7 @@ function resolveEvaluateArgs(
chain: ChainState, chain: ChainState,
): { lastRole: string; lastOutput: EvaluateLastOutput } { ): { lastRole: string; lastOutput: EvaluateLastOutput } {
if (chain.headIsStart) { if (chain.headIsStart) {
return { lastRole: START_ROLE, lastOutput: { [STATUS_KEY]: "_" } }; return { lastRole: START_ROLE, lastOutput: { [STATUS_KEY]: "new" } };
} }
const lastStep = chain.stepsNewestFirst[0]; const lastStep = chain.stepsNewestFirst[0];
@@ -1037,7 +1037,6 @@ function archiveThread(uwf: UwfStore, threadId: ThreadId, _workflow: CasRef, _he
completeThread(uwf.varStore, threadId, "completed"); completeThread(uwf.varStore, threadId, "completed");
} }
// biome-ignore lint/complexity/noExcessiveCognitiveComplexity: orchestration function with inherent branching
export async function cmdThreadResume( export async function cmdThreadResume(
storageRoot: string, storageRoot: string,
threadId: ThreadId, threadId: ThreadId,
@@ -1101,7 +1100,7 @@ export async function cmdThreadResume(
// status === "completed" // status === "completed"
const workflow = loadWorkflowPayload(uwf, workflowHash); const workflow = loadWorkflowPayload(uwf, workflowHash);
const startResult = evaluate(workflow.graph, START_ROLE, {}); const startResult = evaluate(workflow.graph, START_ROLE, { [STATUS_KEY]: "resume" });
if (!startResult.ok) { if (!startResult.ok) {
fail(`failed to evaluate $START: ${startResult.error.message}`); fail(`failed to evaluate $START: ${startResult.error.message}`);
} }
@@ -1113,11 +1112,7 @@ export async function cmdThreadResume(
} }
const startRole = startResult.value.role; const startRole = startResult.value.role;
const completedPromptPrefix = "Previous run completed. Resuming with additional context."; const completedResumePrompt = buildResumePrompt(startResult.value.prompt, supplement);
const completedResumePrompt =
supplement !== null && supplement !== ""
? `${completedPromptPrefix}\n\n${supplement}`
: completedPromptPrefix;
const updatedEntry = { ...entry, status: "idle" as const, completedAt: null }; const updatedEntry = { ...entry, status: "idle" as const, completedAt: null };
setThread(uwf.varStore, threadId, updatedEntry); setThread(uwf.varStore, threadId, updatedEntry);
@@ -6,11 +6,11 @@ describe("Edge prompt template variable resolution", () => {
test("returns error when rendered prompt is empty string", () => { test("returns error when rendered prompt is empty string", () => {
const graph = { const graph = {
$START: { $START: {
_: { role: "classifier", prompt: "{{{userPrompt}}}", location: null }, new: { role: "classifier", prompt: "{{{userPrompt}}}", location: null },
}, },
}; };
const result = evaluate(graph, "$START", {}); const result = evaluate(graph, "$START", { $status: "new" });
expect(result.ok).toBe(false); expect(result.ok).toBe(false);
if (!result.ok) { if (!result.ok) {
@@ -22,11 +22,11 @@ describe("Edge prompt template variable resolution", () => {
test("returns error when rendered prompt is whitespace-only", () => { test("returns error when rendered prompt is whitespace-only", () => {
const graph = { const graph = {
$START: { $START: {
_: { role: "classifier", prompt: " {{{userPrompt}}} ", location: null }, new: { role: "classifier", prompt: " {{{userPrompt}}} ", location: null },
}, },
}; };
const result = evaluate(graph, "$START", {}); const result = evaluate(graph, "$START", { $status: "new" });
expect(result.ok).toBe(false); expect(result.ok).toBe(false);
if (!result.ok) { if (!result.ok) {
@@ -38,11 +38,11 @@ describe("Edge prompt template variable resolution", () => {
test("succeeds when all template variables resolve to non-empty values", () => { test("succeeds when all template variables resolve to non-empty values", () => {
const graph = { const graph = {
$START: { $START: {
_: { role: "classifier", prompt: "{{{userPrompt}}}", location: null }, new: { role: "classifier", prompt: "{{{userPrompt}}}", location: null },
}, },
}; };
const result = evaluate(graph, "$START", { userPrompt: "Fix the bug" }); const result = evaluate(graph, "$START", { $status: "new", userPrompt: "Fix the bug" });
expect(result.ok).toBe(true); expect(result.ok).toBe(true);
if (result.ok) { if (result.ok) {
@@ -53,11 +53,11 @@ describe("Edge prompt template variable resolution", () => {
test("succeeds with static (no-variable) prompt", () => { test("succeeds with static (no-variable) prompt", () => {
const graph = { const graph = {
$START: { $START: {
_: { role: "classifier", prompt: "Classify this input", location: null }, new: { role: "classifier", prompt: "Classify this input", location: null },
}, },
}; };
const result = evaluate(graph, "$START", {}); const result = evaluate(graph, "$START", { $status: "new" });
expect(result.ok).toBe(true); expect(result.ok).toBe(true);
if (result.ok) { if (result.ok) {
@@ -68,11 +68,11 @@ describe("Edge prompt template variable resolution", () => {
test("succeeds when prompt has mix of static text and unresolved variables", () => { test("succeeds when prompt has mix of static text and unresolved variables", () => {
const graph = { const graph = {
$START: { $START: {
_: { role: "classifier", prompt: "Please handle: {{{userPrompt}}}", location: null }, new: { role: "classifier", prompt: "Please handle: {{{userPrompt}}}", location: null },
}, },
}; };
const result = evaluate(graph, "$START", {}); const result = evaluate(graph, "$START", { $status: "new" });
expect(result.ok).toBe(true); expect(result.ok).toBe(true);
if (result.ok) { if (result.ok) {
@@ -83,11 +83,11 @@ describe("Edge prompt template variable resolution", () => {
test("returns error when ALL variables missing and no static text remains", () => { test("returns error when ALL variables missing and no static text remains", () => {
const graph = { const graph = {
$START: { $START: {
_: { role: "classifier", prompt: "{{{a}}}{{{b}}}", location: null }, new: { role: "classifier", prompt: "{{{a}}}{{{b}}}", location: null },
}, },
}; };
const result = evaluate(graph, "$START", {}); const result = evaluate(graph, "$START", { $status: "new" });
expect(result.ok).toBe(false); expect(result.ok).toBe(false);
}); });
+1 -6
View File
@@ -6,10 +6,7 @@ import type { EvaluateResult, Result } from "./types.js";
// Disable HTML escaping — prompts are plain text, not HTML. // Disable HTML escaping — prompts are plain text, not HTML.
mustache.escape = (text: string) => text; mustache.escape = (text: string) => text;
const START_ROLE = "$START";
const SUSPEND_ROLE = "$SUSPEND"; const SUSPEND_ROLE = "$SUSPEND";
// $START is a special entry node with no agent output — it always uses this key.
const START_STATUS = "_";
type LastOutput = Record<string, unknown>; type LastOutput = Record<string, unknown>;
@@ -21,9 +18,7 @@ export function evaluate(
lastOutput: LastOutput, lastOutput: LastOutput,
): Result<EvaluateResult, Error> { ): Result<EvaluateResult, Error> {
let status: string; let status: string;
if (lastRole === START_ROLE) { if (typeof lastOutput[STATUS_KEY] === "string") {
status = START_STATUS;
} else if (typeof lastOutput[STATUS_KEY] === "string") {
status = lastOutput[STATUS_KEY] as string; status = lastOutput[STATUS_KEY] as string;
} else { } else {
return { return {
+4 -13
View File
@@ -97,9 +97,9 @@ function checkGraphStructure(payload: WorkflowPayload, errors: string[]): void {
if (!graphNodes.has("$START")) { if (!graphNodes.has("$START")) {
errors.push("$START must be defined in graph"); errors.push("$START must be defined in graph");
} else { } else {
const startKeys = Object.keys(payload.graph.$START); const startKeys = new Set(Object.keys(payload.graph.$START));
if (startKeys.length !== 1 || startKeys[0] !== "_") { if (!startKeys.has("new") || !startKeys.has("resume")) {
errors.push('$START must have exactly one edge with status "_"'); errors.push('$START must have edges with statuses "new" and "resume"');
} }
} }
@@ -190,22 +190,13 @@ function checkOneOfDiscriminant(
} }
} }
/** Check status-edge consistency for a user role. "_" is reserved for $START and rejected here. */ /** Check status-edge consistency for a user role. */
function checkStatusEdges( function checkStatusEdges(
roleName: string, roleName: string,
graphKeys: Set<string>, graphKeys: Set<string>,
statusSet: Set<string>, statusSet: Set<string>,
errors: string[], errors: string[],
): void { ): void {
if (graphKeys.has("_")) {
errors.push(`role "${roleName}" must use explicit $status keys in graph, not "_"`);
return;
}
if (statusSet.has("_")) {
errors.push(`role "${roleName}" $status enum must use explicit values, not "_"`);
return;
}
const extraKeys = [...graphKeys].filter((k) => !statusSet.has(k)); const extraKeys = [...graphKeys].filter((k) => !statusSet.has(k));
const missingKeys = [...statusSet].filter((k) => !graphKeys.has(k)); const missingKeys = [...statusSet].filter((k) => !graphKeys.has(k));
if (extraKeys.length > 0) { if (extraKeys.length > 0) {
+3 -3
View File
@@ -57,13 +57,13 @@ function isGraph(value: unknown): boolean {
if (!isRecord(value)) { if (!isRecord(value)) {
return false; return false;
} }
return Object.entries(value).every(([node, statusMap]) => { return Object.values(value).every((statusMap) => {
if (!isRecord(statusMap)) { if (!isRecord(statusMap)) {
return false; return false;
} }
return Object.entries(statusMap).every(([status, target]) => { return Object.entries(statusMap).every(([status, target]) => {
// "_" is only valid as a status key for the $START entry node. // "_" is no longer a valid status key anywhere — $START uses "new"/"resume".
if (status === "_" && node !== "$START") { if (status === "_") {
return false; return false;
} }
return isTarget(target); return isTarget(target);
+1 -1
View File
@@ -1,6 +1,6 @@
{ {
"name": "@united-workforce/eval", "name": "@united-workforce/eval",
"version": "0.1.2", "version": "0.1.3",
"private": false, "private": false,
"files": [ "files": [
"src", "src",
+4 -1
View File
@@ -7,12 +7,15 @@ import {
registerRunCommand, registerRunCommand,
} from "./commands/index.js"; } from "./commands/index.js";
// eslint-disable-next-line -- dynamic import for version
const pkg = await import("../package.json", { with: { type: "json" } });
const program = new Command(); const program = new Command();
program program
.name("uwf-eval") .name("uwf-eval")
.description("Evaluate uwf workflow quality with real agents") .description("Evaluate uwf workflow quality with real agents")
.version("0.1.0"); .version(pkg.default.version, "-V, --version");
registerRunCommand(program); registerRunCommand(program);
registerReportCommand(program); registerReportCommand(program);
+1 -1
View File
@@ -1,6 +1,6 @@
{ {
"name": "@united-workforce/util", "name": "@united-workforce/util",
"version": "0.1.0", "version": "0.1.2",
"files": [ "files": [
"src", "src",
"dist", "dist",
-40
View File
@@ -1,40 +0,0 @@
export function generateBootstrapReference(): string {
return `---
name: uwf
description: "United Workforce (uwf) — YAML 状态机工作流引擎。任务涉及 workflow 时加载此 skill。"
tags: [workflow, uwf]
triggers:
- uwf
- workflow
- 工作流
---
# uwf (United Workforce)
YAML 状态机工作流引擎。当用户提到「workflow」「工作流」时,指的是 **uwf workflow**(YAML 定义的状态机),不是 Hermes skill。用 \`uwf\` CLI 操作,不要混淆。
## 首次使用
运行以下命令获取完整用法:
\`\`\`bash
uwf prompt usage # 完整用法文档(所有引用合并)
uwf prompt workflow-authoring # workflow 编写指南(role 定义、graph 路由、schema)
uwf prompt adapter-developing # adapter 开发指南(构建新的 agent adapter)
\`\`\`
## 快速参考
\`\`\`bash
uwf workflow list # 查看已注册 workflow
uwf workflow add <file.yaml> # 注册 workflow
uwf thread start <workflow> -p "prompt" # 创建 thread
uwf thread exec <thread-id> -c 10 # 执行最多 10 步
uwf thread list # 查看所有 thread
\`\`\`
## 示例 workflow
参考项目 \`examples/\` 目录下的 YAML 文件(analyze-topic、debate、solve-issue)。
`;
}
+1 -3
View File
@@ -2,7 +2,6 @@ export { generateActorReference } from "./actor-reference.js";
export { generateAdapterDevelopingReference } from "./adapter-developing-reference.js"; export { generateAdapterDevelopingReference } from "./adapter-developing-reference.js";
export { generateArchitectureReference } from "./architecture-reference.js"; export { generateArchitectureReference } from "./architecture-reference.js";
export { encodeUint64AsCrockford } from "./base32.js"; export { encodeUint64AsCrockford } from "./base32.js";
export { generateBootstrapReference } from "./bootstrap-reference.js";
export { generateCliReference } from "./cli-reference.js"; export { generateCliReference } from "./cli-reference.js";
export { env } from "./env.js"; export { env } from "./env.js";
export type { export type {
@@ -16,7 +15,7 @@ export {
validateFrontmatter, validateFrontmatter,
} from "./frontmatter-markdown/index.js"; } from "./frontmatter-markdown/index.js";
export { createLogger } from "./logger.js"; export { createLogger } from "./logger.js";
export { generateModeratorReference } from "./moderator-reference.js";
export type { export type {
CreateProcessLoggerOptions, CreateProcessLoggerOptions,
ProcessLogFn, ProcessLogFn,
@@ -36,4 +35,3 @@ export { extractUlidTimestamp, generateUlid } from "./ulid.js";
export { generateUsageReference } from "./usage-reference.js"; export { generateUsageReference } from "./usage-reference.js";
export { VERSION } from "./version.js"; export { VERSION } from "./version.js";
export { generateWorkflowAuthoringReference } from "./workflow-authoring-reference.js"; export { generateWorkflowAuthoringReference } from "./workflow-authoring-reference.js";
export { generateYamlReference } from "./yaml-reference.js";
-56
View File
@@ -1,56 +0,0 @@
export function generateModeratorReference(): string {
return `# Moderator Reference
## Overview
The moderator is the workflow engine's routing component. It evaluates the directed graph defined in the workflow YAML to determine the next role (or \`$END\`) after each step — with zero LLM cost.
## Status-Based Routing
The moderator uses **status-based routing**: it inspects the previous step's extracted output (specifically the \`$status\` field) and looks up the corresponding edge in the graph.
### Graph Structure
The graph is a nested map: \`Record<Role | "$START", Record<Status, Target>>\`. Each role maps its possible \`$status\` values to a target with a \`role\` and \`prompt\`:
\`\`\`yaml
graph:
$START:
_: { role: planner, prompt: "Analyze the issue." }
planner:
ready: { role: developer, prompt: "Implement the plan (CAS hash: {{{plan}}})." }
insufficient_info: { role: $END, prompt: "Not enough info." }
developer:
done: { role: reviewer, prompt: "Review branch {{{branch}}} at {{{worktree}}}." }
failed: { role: $END, prompt: "Developer failed: {{{reason}}}." }
reviewer:
approved: { role: tester, prompt: "Run tests on {{{branch}}} at {{{worktree}}}." }
rejected: { role: developer, prompt: "Fix issues: {{{comments}}}." }
\`\`\`
### Routing Algorithm
1. Look up \`graph[lastRole]\` to get the status map for the current role
2. Look up \`statusMap[lastOutput.$status]\` to get the target
3. If target role is \`$END\`, mark thread as completed
4. Otherwise, render the edge prompt (Mustache templates with \`{{{field}}}\` from output) and spawn the next agent
### Edge Prompts and Mustache Templates
Edge prompts use triple-brace Mustache syntax (\`{{{field}}}\`) to interpolate values from the previous step's output into the next agent's task prompt. This passes structured data (branch names, file paths, CAS hashes) between roles without manual wiring.
## Special Nodes
- \`$START\` — entry point; uses status key \`_\` (unconditional) since there is no previous output
- \`$END\` — terminal node; thread completes when reached and is moved to history
## Integration with Steps
Each \`uwf thread exec\` cycle:
1. Moderator reads the thread's head step output
2. Looks up \`graph[lastRole][output.$status]\` to pick the next role
3. If next is \`$END\`, marks thread as completed
4. Otherwise, renders the edge prompt and spawns the agent for the selected role
5. Extract pipeline parses agent output → new step node → append to CAS chain
`;
}
+9 -2
View File
@@ -1,2 +1,9 @@
// This version is kept in sync with package.json during releases. import { readFileSync } from "node:fs";
export const VERSION = "0.1.0"; import { dirname, join } from "node:path";
import { fileURLToPath } from "node:url";
const __dirname = dirname(fileURLToPath(import.meta.url));
const pkg = JSON.parse(readFileSync(join(__dirname, "..", "package.json"), "utf-8")) as {
version: string;
};
export const VERSION = pkg.version;
@@ -40,7 +40,8 @@ roles: # named actors
graph: # status-based routing graph: # status-based routing
$START: $START:
_: { role: planner, prompt: "Analyze the issue." } new: { role: planner, prompt: "Analyze the issue." }
resume: { role: planner, prompt: "Review the previous run output and continue." }
planner: planner:
ready: { role: developer, prompt: "Implement {{{plan}}}." } ready: { role: developer, prompt: "Implement {{{plan}}}." }
failed: { role: $END, prompt: "Failed: {{{error}}}" } failed: { role: $END, prompt: "Failed: {{{error}}}" }
@@ -113,7 +114,7 @@ graph[role][$status] → { role: nextRole, prompt: edgePrompt }
| Node | Purpose | | Node | Purpose |
|------|---------| |------|---------|
| \`$START\` | Entry point — status key is always \`_\` (unconditional) | | \`$START\` | Entry point — status keys \`new\` (first start) and \`resume\` (resuming a completed thread) |
| \`$END\` | Terminal — thread completes and is archived | | \`$END\` | Terminal — thread completes and is archived |
### Edge Prompts ### Edge Prompts
@@ -178,7 +179,7 @@ ocas get <output-hash>
1. Every \`$status\` value in a role's frontmatter has a matching edge in the graph 1. Every \`$status\` value in a role's frontmatter has a matching edge in the graph
2. Every field referenced in edge prompts (\`{{{field}}}\`) exists in the source role's schema 2. Every field referenced in edge prompts (\`{{{field}}}\`) exists in the source role's schema
3. Every role referenced in the graph exists in \`roles\` 3. Every role referenced in the graph exists in \`roles\`
4. \`$START\` has exactly one edge with key \`_\` 4. \`$START\` has edges with keys \`new\` and \`resume\`
5. At least one path leads to \`$END\` 5. At least one path leads to \`$END\`
6. No orphan roles (defined but never routed to) 6. No orphan roles (defined but never routed to)
-82
View File
@@ -1,82 +0,0 @@
export function generateYamlReference(): string {
return `# Workflow YAML Schema Reference
## Top-Level Structure
A workflow YAML file defines the complete workflow specification:
\`\`\`yaml
name: solve-issue # verb-first kebab-case identifier
description: "..." # human-readable description
roles: # named actors in the workflow
planner:
description: "Analyzes issue and outputs a plan"
goal: "You are a planning agent."
capabilities:
- issue-analysis
- planning
procedure: |
1. Read the issue
2. Produce a test spec
output: "Output the plan summary. Set $status to ready or insufficient_info."
frontmatter: # JSON Schema for structured output (drives routing)
oneOf:
- properties:
$status: { const: ready }
plan: { type: string }
required: [$status, plan]
- properties:
$status: { const: insufficient_info }
required: [$status]
graph: # status-based routing (nested map)
$START:
_: { role: planner, prompt: "Analyze the issue." }
planner:
ready: { role: developer, prompt: "Implement plan {{{plan}}}." }
insufficient_info: { role: $END, prompt: "Not enough info." }
\`\`\`
## roles
Each role defines an actor in the workflow:
| Field | Type | Description |
|-------|------|-------------|
| \`description\` | string | Short description of the role's purpose |
| \`goal\` | string | System-level goal statement for the agent |
| \`capabilities\` | string[] | Tags describing what the role can do |
| \`procedure\` | string | Step-by-step instructions for the agent |
| \`output\` | string | Description of expected output format |
| \`frontmatter\` | JSON Schema | Defines the structured output the agent must produce |
### frontmatter
The \`frontmatter\` field is a standard JSON Schema object. The extract pipeline validates agent output against it. Key conventions:
- \`$status\` field drives routing decisions in the graph
- Use \`const\` or \`enum\` to constrain status values
- Use \`oneOf\` to define multiple valid output shapes (one per status)
- All \`required\` fields must appear in the agent's frontmatter output
## graph
The graph is a nested map defining status-based routing:
\`\`\`
Record<Role | "$START", Record<Status, { role: string, prompt: string }>>
\`\`\`
| Level | Key | Value |
|-------|-----|-------|
| Outer | Role name or \`$START\` | Status map for that role |
| Inner | \`$status\` value (or \`_\` for unconditional) | Target: \`{ role, prompt }\` |
### Special Nodes
- \`$START\` — entry point; uses status key \`_\` (unconditional, no previous output)
- \`$END\` — terminal node; thread completes when reached
### Edge Prompts
Prompts use triple-brace Mustache templates (\`{{{field}}}\`) to interpolate values from the previous step's output. Example: \`"Implement plan {{{plan}}} in repo {{{repoPath}}}."\`
`;
}
+4 -1
View File
@@ -21,9 +21,12 @@ graph:
role: package-metadata role: package-metadata
prompt: Biome setup failed ({{{reason}}}), but continue. Standardize package metadata for repo at {{{repoPath}}}. prompt: Biome setup failed ({{{reason}}}), but continue. Standardize package metadata for repo at {{{repoPath}}}.
$START: $START:
_: new:
role: workspace role: workspace
prompt: Set up bun workspace structure for repo at {{{repoPath}}}. prompt: Set up bun workspace structure for repo at {{{repoPath}}}.
resume:
role: workspace
prompt: Review the previous run output and continue setting up the bun workspace structure for repo at {{{repoPath}}}.
release: release:
done: done:
role: testing role: testing
+4 -1
View File
@@ -283,9 +283,12 @@ roles:
- error - error
graph: graph:
$START: $START:
_: new:
role: planner role: planner
prompt: Analyze the issue and produce an implementation plan. prompt: Analyze the issue and produce an implementation plan.
resume:
role: planner
prompt: Review the previous run output and continue the work.
planner: planner:
insufficient_info: insufficient_info:
role: $SUSPEND role: $SUSPEND