agent-afk 3.110.1 → 3.111.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (5) hide show
  1. package/README.md +10 -16
  2. package/dist/cli.mjs +358 -358
  3. package/dist/index.mjs +129 -129
  4. package/dist/telegram.mjs +148 -148
  5. package/package.json +1 -1
package/README.md CHANGED
@@ -1,8 +1,8 @@
1
1
  # Agent AFK
2
2
 
3
- > Delegate the work. Keep the judgment.
3
+ > Run coding agents while you’re AFK.
4
4
  >
5
- > **Agent AFK** is a local-first control plane for long-running AI agent work. Hand off multi-session, headless tasks to parallel agents — then stay in control with traces you can read (`afk trace show`), resumable sessions, verification passes before anything ships, and a phone ping the moment work lands in a terminal state. Bring your own model.
5
+ > Start a run in your terminal, walk away, and get pinged when Agent AFK finishes or needs you. Every step is saved as a readable trace, so you stay in control before anything ships.
6
6
 
7
7
  [![npm version](https://img.shields.io/npm/v/agent-afk.svg)](https://www.npmjs.com/package/agent-afk)
8
8
  [![Node](https://img.shields.io/node/v/agent-afk.svg)](https://nodejs.org/)
@@ -26,7 +26,7 @@ afk chat "hello"
26
26
 
27
27
  ## What you can do with it
28
28
 
29
- - **Chat from your terminal** — `afk chat "..."` for one-shot, `afk i` for a REPL with full tool access (Bash, file ops, web fetch, grep/glob, subagents).
29
+ - **Chat from your terminal** — `afk chat "..."` for one-shot, `afk` for a REPL with full tool access (Bash, file ops, web fetch, grep/glob, subagents).
30
30
  - **Hand long work off to a daemon** — `afk daemon` runs headless. Pair it with `send_telegram` and you get pings on your phone when work lands in a terminal state.
31
31
  - **Read the trace** — every run writes an append-only record of what the agent did. `afk trace show` prints it back as a human-readable receipt — tool calls, gate decisions, subagent lifecycles, cost — so you can audit a run without reaching for `jq`.
32
32
  - **Message Claude from Telegram** — `afk telegram setup` walks you through bot token + allowlist. After that you have a private chat surface backed by the same session manager as the REPL.
@@ -40,8 +40,8 @@ afk chat "hello"
40
40
 
41
41
  | Command | Surface |
42
42
  |---|---|
43
- | `afk chat "..."` | One-shot turn — pipe-friendly, scripts well |
44
- | `afk i` (alias of `afk interactive`) | REPL with slash commands, streaming, plan mode, image paste |
43
+ | `afk chat "..."` | One-shot (fire & forget) turn — pipe-friendly, scripts well |
44
+ | `afk` (alias of `afk interactive`) | REPL with slash commands, streaming, plan mode, image paste |
45
45
  | `afk daemon` | Long-running headless agent, cron-friendly |
46
46
  | `afk telegram start` | Telegram bot — same tools, same memory, on your phone |
47
47
 
@@ -51,12 +51,6 @@ afk chat "hello"
51
51
 
52
52
  **Local-first, no phone-home.** There is no analytics or remote telemetry — Agent AFK never sends your prompts, code, or usage anywhere except directly to the model provider you configure. What telemetry exists is local JSONL under `~/.afk/` that you can read or delete.
53
53
 
54
- Minimum viable config is one env var:
55
-
56
- ```bash
57
- ANTHROPIC_API_KEY=sk-ant-...
58
- ```
59
-
60
54
  Optional, in order of usefulness:
61
55
 
62
56
  ```bash
@@ -80,9 +74,9 @@ AFK_MAX_BUDGET_USD=5.00
80
74
  Default is `sonnet`. Override per-call with `--model`:
81
75
 
82
76
  ```bash
83
- afk chat "explain this stack trace" --model opus
84
- afk i --model haiku
85
- afk chat "refactor this" --model codex
77
+ afk chat "explain this stack trace" --model opus_1m
78
+ afk --model sonnet_1m
79
+ afk chat "refactor this" --model gpt-5.5
86
80
  ```
87
81
 
88
82
  | Model | Best for |
@@ -90,7 +84,7 @@ afk chat "refactor this" --model codex
90
84
  | `fable` | Most capable — Claude Fable 5 (Mythos-class), hardest reasoning + long-horizon agentic work (1M context) |
91
85
  | `opus` | Complex reasoning, multi-step planning, long contexts |
92
86
  | `sonnet` | Day-to-day default — balanced speed and capability |
93
- | `haiku` | Fast, cheap, one-shots |
87
+ | `haiku` | Fast, cheap, meh... |
94
88
 
95
89
  ## Useful commands
96
90
 
@@ -107,7 +101,7 @@ Aliases: `afk c` → `chat`, `afk i` → `interactive`, `afk s` → `status`.
107
101
 
108
102
  ## A note on permissions
109
103
 
110
- `afk` runs with **bypass permissions** by default: no per-tool prompts. Claude can run bash, read and write files, fetch URLs, and call MCP servers without asking each time. This is intentional — `afk` is built for unattended work, where a permission prompt with no human in front of it is just a wedged session.
104
+ `afk` runs with **full permissions** by default: no per-tool prompts. Claude can run bash, read and write files, fetch URLs, and call MCP servers without asking each time. This is intentional — `afk` is built for unattended work, where a permission prompt with no human in front of it is just a wedged session.
111
105
 
112
106
  Use `afk` on a machine and account you trust. Override per-session with `--permission-mode` if you want stricter behavior.
113
107