agent-afk 5.232.0 → 5.232.2

package/README.md

@@ -19,23 +19,22 @@ Open source. Runs locally. Works with any model provider, including local models
 
 > ⭐ **Like the idea of an agent you can walk away from? [Star the repo](https://github.com/griffinwork40/agent-afk/stargazers)** -- it's the fastest way to help other people find it.
 
-## Install
+## Quick Start
 
 ```bash
-npm install -g agent-afk
+npm install -g agent-afk    # Node ≥ 22 required
+afk login                   # authenticate (auto-detects Claude Code / Codex creds)
+afk doctor                  # verify everything works
+afk chat "hello"             # first real conversation
 ```
 
-Requires Node ≥ 22.
+That's it. You're in.
 
-Smoke test:
+**Try without installing:** `npx agent-afk chat "hello"` runs a one-shot turn with zero global install.
 
-```bash
-afk --version    # confirm the install (works before login)
-afk doctor       # environment self-check
-afk chat "hello"
-```
+**Already using Claude Code or Codex?** `afk login` will detect your existing credentials automatically. Run `afk migrate` to import your plugins, skills, and MCP servers too -- it live-reads the source tool's dirs, so anything you install there keeps showing up in AFK with no re-run.
 
-**Already using Claude Code or Codex?** Run `afk migrate` — it imports your existing plugins, skills, and MCP servers. It doesn't copy files; it live-reads the source tool's dirs, so anything you install there keeps showing up in AFK with no re-run.
+> 📖 **Full documentation at [docs.agentafk.com](https://docs.agentafk.com)** -- quickstart, configuration, model setup, surfaces, skills, and SDK reference.
 
 ## What you can do with it
 
@@ -231,7 +230,7 @@ With containment on (`default`), a file tool (read/write/edit/list/glob/grep) ta
 
 ## Troubleshooting
 
-**`invalid x-api-key` / `ANTHROPIC_API_KEY not found`** — run `afk doctor`. Confirm the key is set in your shell or in `~/.afk/config/afk.env`.
+**`invalid x-api-key` / `ANTHROPIC_API_KEY not found`** — run `afk login` to authenticate interactively, or set the key manually: `export ANTHROPIC_API_KEY=sk-ant-...` (or add it to `~/.afk/config/afk.env`). Run `afk doctor` to confirm.
 
 **`Cannot send message: session is closed`** — the session timed out or was closed. Start a new one (`afk i` or a fresh `afk chat`).
 

package/dist/agent/branch/branch-prune.d.ts

@@ -0,0 +1,36 @@
+export type ExecForBranchPrune = (file: string, args: string[], opts?: {
+    cwd?: string;
+}) => Promise<{
+    stdout: string;
+    stderr: string;
+}>;
+export type PrState = 'open' | 'merged' | 'closed' | 'none';
+export type BranchVerdict = 'prune' | 'keep' | 'error';
+export interface BranchCandidate {
+    remoteBranch: string;
+    shortName: string;
+    prState: PrState;
+    commitsAhead: number;
+    verdict: BranchVerdict;
+    reason: string;
+    errorDetail?: string;
+}
+export interface BranchPruneResult {
+    candidates: BranchCandidate[];
+    deleted: string[];
+    warnings: string[];
+    dryRun: boolean;
+}
+export interface BranchPruneOptions {
+    execFn: ExecForBranchPrune;
+    remote?: string;
+    baseBranch?: string;
+    branchPrefix?: string;
+    dryRun?: boolean;
+    cwd?: string;
+}
+export declare function parseGhPrState(ghOutput: string): PrState;
+export declare function parseCommitsAhead(revListOutput: string): number;
+export declare function classify(prState: PrState, commitsAhead: number): Pick<BranchCandidate, 'verdict' | 'reason'>;
+export declare function listRemoteBranches(execFn: ExecForBranchPrune, remote: string, branchPrefix: string, cwd: string): Promise<string[]>;
+export declare function runBranchPrune(options: BranchPruneOptions): Promise<BranchPruneResult>;

package/dist/bundled-plugins/awa-bundled/skills/fix-pr/SKILL.md

@@ -74,7 +74,7 @@ This phase validates each spec item independently before any writes begin. The g
 
 **Fast-path:** if the numbered spec contains exactly 1 item, skip this phase entirely -- dispatch directly to Phase 3 with the raw spec item. The parallel wave's value is proportional to item count; for a single item the overhead exceeds the benefit.
 
-**For 2+ spec items:** dispatch N read-only sub-agents in parallel via the `agent` tool (one call per spec item in the same tool-use turn -- pure fan-out). Do NOT use `compose` for this wave: compose passes `agent_type` as a display label only -- it does NOT resolve named agents through the agent registry, so tool-surface restrictions (e.g. research-agent's read-only enforcement) are silently lost. Until compose wires registry resolution (see #2000), only the `agent` tool enforces named-agent permissions. Each probe:
+**For 2+ spec items:** dispatch N read-only sub-agents in parallel via the `agent` tool (one call per spec item in the same tool-use turn -- pure fan-out). `compose` now supports `agent_type`, `cwd`, `readRoots`, and `writeRoots` per node (so the read-only and worktree-scoping guarantees are NOT silently lost when using compose), but the `agent` fan-out is preferred here: it gives simpler rate-limiting control (see the rate note below) and each probe is truly independent with no DAG dependencies to model. If DAG dependency tracking between probes matters for a future extension, compose is a valid alternative -- just wire `agent_type: "research-agent"` and `cwd: <worktree path>` on each node. Each probe:
 
 - **agent_type:** `research-agent` (mechanically read-only -- cannot write files, run bash, or commit).
 - **cwd:** `<worktree path>` (so file reads target the PR branch, not the main working tree).

package/dist/cli/commands/branch.d.ts

@@ -0,0 +1,2 @@
+import { Command } from 'commander';
+export declare function registerBranchCommand(program: Command): void;