docs: add README.md to all 8 packages

Each README includes: package description, install instructions, usage example, API overview, and (for templates) role/moderator flow. Packages documented: - @uncaged/workflow (core) - @uncaged/cli-workflow (CLI) - @uncaged/workflow-agent-cursor - @uncaged/workflow-agent-hermes - @uncaged/workflow-agent-llm - @uncaged/workflow-template-develop - @uncaged/workflow-template-solve-issue - @uncaged/workflow-util-agent
2026-05-08 10:17:40 +08:00
parent f81e2a8aac
commit a5c09adae6
8 changed files with 352 additions and 0 deletions
@@ -0,0 +1,76 @@
+# @uncaged/cli-workflow
+
+Command-line interface for the Uncaged workflow engine (`uncaged-workflow`).
+
+The CLI reads and writes the workflow registry, starts and inspects threads, manages CAS blobs, and prints agent-oriented reference docs via `skill`. It uses the same storage layout as `@uncaged/workflow` (default `~/.uncaged/workflow`).
+
+## Install
+
+```bash
+bun add @uncaged/cli-workflow
+```
+
+In this monorepo: `"@uncaged/cli-workflow": "workspace:*"`. Depends on `"@uncaged/workflow": "workspace:*"`.
+
+## Usage
+
+```bash
+uncaged-workflow workflow list
+uncaged-workflow run <name> --prompt "Your task"
+uncaged-workflow thread show <id>
+uncaged-workflow skill
+```
+
+Invoking the CLI with no command (or from this repo: `bun packages/cli-workflow/src/cli.ts`) prints:
+
+```
+uncaged-workflow — workflow engine CLI
+
+Workflow registry:
+  workflow add <name> <file.esm.js> [--types <path>]  Register a workflow bundle in the registry
+  workflow list                                       List all registered workflows
+  workflow show <name>                                Show details of a registered workflow
+  workflow rm <name>                                  Remove a workflow from the registry
+  workflow history <name>                             Show version history of a workflow
+  workflow rollback <name> [hash]                     Rollback a workflow to a previous version
+
+Thread execution:
+  thread run <name> [--prompt <text>] [--max-rounds N]          Start a new thread executing a workflow
+  thread list [name]                                            List threads, optionally filtered by workflow name
+  thread show <id>                                              Show thread details and state
+  thread rm <id>                                                Remove a thread
+  thread fork <thread-id> [--from-role <role>]                  Fork a thread, optionally from a specific role
+  thread ps                                                     List running threads
+  thread kill <thread-id>                                       Kill a running thread
+  thread live <thread-id> | --latest [--debug] [--role <name>]  Attach to a thread and stream output live
+  thread pause <thread-id>                                      Pause a running thread
+  thread resume <thread-id>                                     Resume a paused thread
+
+Content-addressable storage:
+  cas get <hash>     Retrieve content by hash from CAS
+  cas put <content>  Store content in CAS, prints hash
+  cas list           List all hashes in CAS
+  cas rm <hash>      Remove a CAS entry by hash
+  cas gc             Garbage-collect unreferenced CAS entries
+
+Development:
+  init workspace <name>  Initialize a new workflow workspace
+  init template <name>   Initialize a new workflow template
+
+Shortcuts:
+  run <name> [...]  → thread run
+  live <id> [...]   → thread live
+
+Reference:
+  skill [topic]  Agent-consumable docs (cli, develop, author)
+
+Use <command> --help for subcommand details.
+
+Environment variables:
+  WORKFLOW_STORAGE_ROOT              Override storage directory (default: ~/.uncaged/workflow)
+  UNCAGED_WORKFLOW_STORAGE_ROOT      Internal override (takes priority over WORKFLOW_STORAGE_ROOT)
+```
+
+## API overview
+
+This package is bin-only; programmatic use is via `@uncaged/workflow`. Entry: `src/cli.ts` → `runCli` in `src/cli-dispatch.js`.
@@ -0,0 +1,36 @@
+# @uncaged/workflow-agent-cursor
+
+`AgentFn` adapter that runs the `cursor-agent` CLI against a workspace path derived from the thread.
+
+The agent builds a full prompt (system + task + step history via `@uncaged/workflow-util-agent`), extracts the absolute workspace path with your `extract` + Zod schema, then spawns `cursor-agent` with `--workspace`, model, and non-interactive flags.
+
+## Install
+
+```bash
+bun add @uncaged/workflow-agent-cursor @uncaged/workflow @uncaged/workflow-util-agent zod
+```
+
+In this monorepo: `"@uncaged/workflow-agent-cursor": "workspace:*"` plus `workspace:*` for `@uncaged/workflow` and `@uncaged/workflow-util-agent`.
+
+## Usage
+
+```typescript
+import { createCursorAgent } from "@uncaged/workflow-agent-cursor";
+
+const agent = createCursorAgent({
+  model: null, // null → "auto"
+  timeout: 0, // ms; 0 = no limit (spawnCli timeout disabled)
+  extract: myExtractFn,
+});
+```
+
+## API overview
+
+| Export | Description |
+|--------|-------------|
+| `createCursorAgent(config)` | Returns `AgentFn` that runs `cursor-agent` with `buildAgentPrompt(ctx)` |
+| `CursorAgentConfig` | `model`, `timeout`, `extract` (must supply workspace path) |
+| `validateCursorAgentConfig` | Config validation result |
+| `buildAgentPrompt` | Re-exported from `@uncaged/workflow-util-agent` |
+
+Requires `cursor-agent` on `PATH` at runtime.
@@ -0,0 +1,35 @@
+# @uncaged/workflow-agent-hermes
+
+`AgentFn` adapter that runs the `hermes` CLI in non-interactive `chat` mode (Nerve-style flags: `-q`, `--yolo`, `--quiet`, bounded `--max-turns`).
+
+The agent composes the same thread-aware prompt as other CLI-backed agents via `buildAgentPrompt` from `@uncaged/workflow-util-agent`, then spawns `hermes` and returns stdout on success.
+
+## Install
+
+```bash
+bun add @uncaged/workflow-agent-hermes @uncaged/workflow @uncaged/workflow-util-agent
+```
+
+In this monorepo: use `workspace:*` for all three `@uncaged/*` packages.
+
+## Usage
+
+```typescript
+import { createHermesAgent } from "@uncaged/workflow-agent-hermes";
+
+const agent = createHermesAgent({
+  model: "your-model", // or null to omit --model
+  timeout: 600_000, // ms, or null for no timeout
+});
+```
+
+## API overview
+
+| Export | Description |
+|--------|-------------|
+| `createHermesAgent(config)` | Returns `AgentFn` wrapping `hermes chat -q ...` |
+| `HermesAgentConfig` | `model`, `timeout` |
+| `validateHermesAgentConfig` | Config validation result |
+| `buildAgentPrompt` | Re-exported from `@uncaged/workflow-util-agent` |
+
+Requires `hermes` on `PATH` at runtime.
@@ -0,0 +1,34 @@
+# @uncaged/workflow-agent-llm
+
+`AgentFn` adapter that calls an OpenAI-compatible `POST /chat/completions` endpoint using `@uncaged/workflow`’s `LlmProvider` (base URL, API key, model).
+
+Single-turn: system text is the current role’s `systemPrompt`, user text is the thread’s initial prompt (`ctx.start.content`). Errors from HTTP, JSON, or empty choices are thrown as `Error` with a JSON payload string.
+
+## Install
+
+```bash
+bun add @uncaged/workflow-agent-llm @uncaged/workflow
+```
+
+In this monorepo: `"@uncaged/workflow-agent-llm": "workspace:*"`, `"@uncaged/workflow": "workspace:*"`.
+
+## Usage
+
+```typescript
+import { createLlmAdapter } from "@uncaged/workflow-agent-llm";
+
+const agent = createLlmAdapter({
+  baseUrl: "https://api.openai.com/v1",
+  apiKey: process.env.OPENAI_API_KEY!,
+  model: "gpt-4.1-mini",
+});
+```
+
+## API overview
+
+| Export | Description |
+|--------|-------------|
+| `createLlmAdapter(provider)` | `LlmProvider` → `AgentFn` |
+| `chatCompletionText({ provider, messages })` | Low-level `Result<string, LlmChatError>` helper |
+| `LlmMessage` | `{ role: "system" \| "user" \| "assistant"; content: string }` |
+| `LlmChatError` | Discriminated error kinds for failed completions |
@@ -0,0 +1,53 @@
+# @uncaged/workflow-template-develop
+
+Reference **develop** workflow template: plan phases, implement in a loop, review, test, then commit.
+
+Export a `WorkflowDefinition` and `createDevelopRun` so a host can bind agents/LLM and run the same graph the bundled `.esm.js` would use. Use `buildDevelopDescriptor()` when assembling `descriptor` metadata for a bundle.
+
+## Install
+
+```bash
+bun add @uncaged/workflow-template-develop @uncaged/workflow zod
+```
+
+In this monorepo: `workspace:*` for `@uncaged/workflow-template-develop` and `@uncaged/workflow`.
+
+## Usage
+
+```typescript
+import { createDevelopRun, developWorkflowDefinition } from "@uncaged/workflow-template-develop";
+
+const run = createDevelopRun(binding, extract, llmProvider);
+// run(...) executes the develop moderator graph with your AgentBinding
+```
+
+## Roles
+
+| Role | Purpose |
+|------|---------|
+| **planner** | Break work into ordered phases (hashes) |
+| **coder** | Implement current phase; repeats until phases complete or limits hit |
+| **reviewer** | Code review gate (`approved` vs send back to coder) |
+| **tester** | Verify via tests/build/lint (`passed` vs send back to coder) |
+| **committer** | Final commit step |
+
+Also exported: role factories/meta schemas (`plannerRole`, `coderRole`, …), `DevelopMeta`, `developRoles`.
+
+## Moderator flow
+
+1. **Start** → `planner`  
+2. After **planner** → `coder`  
+3. After **coder** → if all planned phases are done (or last phase completed) → `reviewer`; else `coder` again, until `maxRounds` then `END`  
+4. After **reviewer** → if approved → `tester`; else `coder` (or `END` if out of rounds)  
+5. After **tester** → if passed → `committer`; else `coder` (or `END` if out of rounds)  
+6. After **committer** → `END`
+
+## API overview
+
+| Export | Description |
+|--------|-------------|
+| `createDevelopRun` | `createWorkflow(developWorkflowDefinition, …)` factory |
+| `developWorkflowDefinition` | `description`, `roles`, `developModerator` |
+| `developModerator` | `Moderator<DevelopMeta>` |
+| `buildDevelopDescriptor` | `buildDescriptor({ … })` for bundle metadata |
+| `DEVELOP_WORKFLOW_DESCRIPTION` | Human-readable one-liner |
@@ -0,0 +1,48 @@
+# @uncaged/workflow-template-solve-issue
+
+Reference **solve-issue** workflow template: prepare a repo, delegate implementation to the **develop** workflow, then submit (e.g. open a PR).
+
+`createSolveIssueRun` wires the `developer` role to `workflowAsAgent("develop")` by default; `binding.overrides.developer` wins if you pass one (for tests or custom hosts).
+
+## Install
+
+```bash
+bun add @uncaged/workflow-template-solve-issue @uncaged/workflow zod
+```
+
+In this monorepo: `workspace:*` for this package and `@uncaged/workflow`.
+
+## Usage
+
+```typescript
+import { createSolveIssueRun, solveIssueWorkflowDefinition } from "@uncaged/workflow-template-solve-issue";
+
+const run = createSolveIssueRun(binding, extract, llmProvider);
+```
+
+## Roles
+
+| Role | Purpose |
+|------|---------|
+| **preparer** | Set up context / repo state for the issue |
+| **developer** | Implementation; default runs the registered `develop` workflow as a sub-agent |
+| **submitter** | Finalize and submit the outcome (e.g. PR) |
+
+Also exported: `preparerRole`, `developerRole`, `submitterRole` and their Zod meta schemas, `SolveIssueMeta`, `solveIssueRoles`.
+
+## Moderator flow
+
+1. **Start** → `preparer`  
+2. After **preparer** → `developer`  
+3. After **developer** → `submitter`  
+4. After **submitter** → `END`
+
+## API overview
+
+| Export | Description |
+|--------|-------------|
+| `createSolveIssueRun` | Merges `developer` override with `workflowAsAgent("develop")`, then `createWorkflow` |
+| `solveIssueWorkflowDefinition` | `description`, `roles`, `solveIssueModerator` |
+| `solveIssueModerator` | Linear `Moderator<SolveIssueMeta>` |
+| `buildSolveIssueDescriptor` | Descriptor helper for bundles |
+| `SOLVE_ISSUE_WORKFLOW_DESCRIPTION` | Human-readable one-liner |
@@ -0,0 +1,34 @@
+# @uncaged/workflow-util-agent
+
+Shared helpers for CLI-backed workflow agents: assemble prompts from thread context and spawn subprocesses with timeouts.
+
+Used by `@uncaged/workflow-agent-cursor` and `@uncaged/workflow-agent-hermes`. Depends on `@uncaged/workflow` for CAS reads (`getContentMerklePayload`) and `Result` typing.
+
+## Install
+
+```bash
+bun add @uncaged/workflow-util-agent @uncaged/workflow
+```
+
+In this monorepo: `workspace:*` for both packages.
+
+## Usage
+
+```typescript
+import { buildAgentPrompt, spawnCli } from "@uncaged/workflow-util-agent";
+
+const prompt = await buildAgentPrompt(agentContext);
+const result = await spawnCli("my-cli", ["--json"], { cwd: "/tmp", timeoutMs: 60_000 });
+if (!result.ok) { /* handle SpawnCliError */ }
+const stdout = result.value;
+```
+
+## API overview
+
+| Export | Description |
+|--------|-------------|
+| `buildAgentPrompt(ctx)` | System prompt + task + prior step summaries + latest body from CAS; appends `uncaged-workflow thread <id>` tool hint |
+| `spawnCli(cmd, args, { cwd, timeoutMs })` | `Promise<Result<string, SpawnCliError>>`; captures stdout, non-zero exit and spawn failures as `err` |
+| `SpawnCliConfig` | `cwd: string \| null`, `timeoutMs: number \| null` |
+| `SpawnCliError` | `non_zero_exit` \| `timeout` \| `spawn_failed` |
+| `SpawnCliResult` | Alias for `Result<string, SpawnCliError>` |
@@ -0,0 +1,36 @@
+# @uncaged/workflow
+
+Core workflow engine: registry, CAS, thread execution, bundle validation, and role/workflow types.
+
+This package implements the three-phase engine loop that runs single-file ESM workflow bundles (each exports `run` and `descriptor`). It persists threads under `~/.uncaged/workflow/` by default and hashes bundles with XXH64 (Crockford Base32). See the repo root [README](../../README.md) for workflow, bundle, thread, role, and registry concepts.
+
+## Install
+
+```bash
+bun add @uncaged/workflow zod
+```
+
+In this monorepo, depend with `"@uncaged/workflow": "workspace:*"`. `zod` is a peer dependency (used by bundle/shape validation at the integration boundary).
+
+## Usage
+
+```typescript
+import { createWorkflow, readWorkflowRegistry, executeThread } from "@uncaged/workflow";
+// Wire a WorkflowDefinition + AgentBinding + extract + optional LlmProvider into createWorkflow,
+// then run the returned WorkflowFn inside your host (or use executeThread for disk-backed runs).
+```
+
+## API overview
+
+| Area | Exports (representative) |
+|------|--------------------------|
+| **Types** | `WorkflowDefinition`, `WorkflowFn`, `AgentFn`, `AgentBinding`, `Moderator`, `RoleDefinition`, `ThreadContext`, `LlmProvider`, `Result` shape via `ok` / `err`, `START` / `END` |
+| **Bundle** | `buildDescriptor`, `extractBundleExports`, `validateWorkflowBundle`, `validateWorkflowDescriptor`, `WorkflowDescriptor`, `WorkflowRoleDescriptor` |
+| **Registry** | `readWorkflowRegistry`, `writeWorkflowRegistry`, `registerWorkflowVersion`, `workflowRegistryPath`, YAML helpers |
+| **CAS** | `createCasStore`, `createThreadCas`, Merkle helpers (`putStepMerkleNode`, `getContentMerklePayload`, …), `hashWorkflowBundleBytes` |
+| **Engine** | `createWorkflow`, `executeThread`, `parseThreadDataJsonl`, fork helpers, `garbageCollectCas` |
+| **Extract / LLM tools** | `llmExtract`, `reactExtract`, `createExtract`, `getExtractProvider` |
+| **Agent bridge** | `workflowAsAgent` — expose a registered workflow as an agent-backed role |
+| **Utilities** | `createLogger`, ULID / Crockford Base32 codecs, `getDefaultWorkflowStorageRoot`, paths |
+
+Full surface is re-exported from `src/index.ts`.