pikiloom 0.4.5 → 0.4.6
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.
- package/README.md +24 -20
- package/package.json +1 -1
package/README.md
CHANGED
|
@@ -6,7 +6,7 @@
|
|
|
6
6
|
|
|
7
7
|
##### *The open Agent orchestrator for the era when creators no longer need to read code.*
|
|
8
8
|
|
|
9
|
-
*Plug in any agent (Claude · Codex · Gemini · Hermes · …), any model (Claude · GPT · Gemini · DeepSeek · Doubao · MiMo · MiniMax · OpenRouter · or any
|
|
9
|
+
*Plug in any agent (Claude · Codex · Gemini · Hermes · …), any model (Claude · GPT · Gemini · DeepSeek · Doubao · MiMo · MiniMax · Qwen · OpenRouter · local Ollama / mlx-lm · or any OpenAI-compatible proxy), and any tool (Skills · MCP · CLI) — then drive a **swarm of them in parallel** from whichever terminal is closest: an IM, the Web Dashboard, or the CLI. pikiloom is built with pikiloom.*
|
|
10
10
|
|
|
11
11
|
```bash
|
|
12
12
|
npx pikiloom@latest
|
|
@@ -79,42 +79,46 @@ This is the shape that matters: one creator, with a swarm of AI agents at their
|
|
|
79
79
|
|
|
80
80
|
## See It in Action
|
|
81
81
|
|
|
82
|
-
> **
|
|
82
|
+
> **A Swarm, in Parallel** — The whole point: run **N agents across N panes at once**. Every pane is an independent agent stream on its own session and workspace. Dispatch a task to each and watch them all work simultaneously — then fan out from 1 → 2 → 3 → 6 panes as you scale.
|
|
83
|
+
|
|
84
|
+
<p align="center"><img src="docs/promo-parallel.gif" alt="Multiple agents working in parallel across panes — dispatch to all, watch them stream at once" width="820"></p>
|
|
83
85
|
|
|
84
|
-
<p align="center"><img src="docs/promo-
|
|
86
|
+
<p align="center"><img src="docs/promo-dashboard-workspace.png" alt="Three agents running in parallel — three panes, three live sessions" width="820"></p>
|
|
85
87
|
|
|
86
|
-
> **
|
|
88
|
+
> **Real-world Task** — Ask pikiloom to gather and summarize today's AI news; the agent reads, writes, and ships the results back through Telegram, all controlled from your phone.
|
|
87
89
|
|
|
88
|
-
<p align="center"><img src="docs/promo-
|
|
90
|
+
<p align="center"><img src="docs/promo-demo.gif" alt="Telegram demo: ask, the agent works locally, the result returns to chat" width="780"></p>
|
|
89
91
|
|
|
90
92
|
<details>
|
|
91
|
-
<summary><b>More:
|
|
93
|
+
<summary><b>More: Six-pane grid · Switch agents mid-session · Standard/Extra mode · IM · Agents & Models · Extensions · System</b></summary>
|
|
92
94
|
|
|
93
|
-
>
|
|
95
|
+
> **Six in parallel** — Six independent agent streams on one screen. Pick the layout (1 / 2 / 3 / 6) that matches how wide you want to fan out; each pane keeps its own agent, model, and session history. Light/dark theme and full EN/中文 throughout.
|
|
94
96
|
|
|
95
|
-
<img src="docs/promo-
|
|
97
|
+
<img src="docs/promo-dashboard-six.png" alt="Six-pane parallel grid — six agent streams at once" width="820">
|
|
96
98
|
|
|
97
|
-
> **
|
|
99
|
+
> **Switch the Agent mid-session** — Swap the "brain" on the fly — Claude Code · Codex · Gemini · Hermes — without leaving the workspace; the same conversation history follows you to the next agent.
|
|
98
100
|
|
|
99
|
-
<img src="docs/promo-dashboard-
|
|
101
|
+
<img src="docs/promo-dashboard-agent-switch.png" alt="Per-session agent switcher" width="820">
|
|
100
102
|
|
|
101
|
-
> **
|
|
103
|
+
> **Standard / Extra access mode** — Per agent, choose which billing pool Claude turns land on: **Standard** (interactive, counted inside your Pro/Max subscription) or **Extra** (headless `claude -p`, billed against the separate Agent SDK credit pool). Plus per-session model and reasoning-effort selectors.
|
|
102
104
|
|
|
103
|
-
<img src="docs/promo-dashboard-
|
|
105
|
+
<img src="docs/promo-dashboard-agent-mode.png" alt="Standard / Extra access mode toggle" width="820">
|
|
104
106
|
|
|
105
|
-
> **
|
|
107
|
+
> **IM Access** — Check and configure connection statuses for Telegram, Feishu, WeChat, Slack, Discord, DingTalk, and WeCom.
|
|
106
108
|
|
|
107
|
-
|
|
109
|
+
<img src="docs/promo-dashboard-im.png" alt="IM Access" width="820">
|
|
108
110
|
|
|
109
|
-
|
|
111
|
+
> **Agents & Models** — Manage installed agent CLIs, set your default agent, and bind a secure Providers + Profiles vault (Claude · GPT · Gemini · DeepSeek · Doubao · MiMo · MiniMax · Qwen · OpenRouter · any OpenAI-compatible proxy), validated against the read-only `models.dev` catalog and injected per-agent at spawn time. Local backends (Ollama, mlx-lm on Apple Silicon) attach the moment they're detected.
|
|
110
112
|
|
|
111
|
-
|
|
113
|
+
<img src="docs/promo-dashboard-agents.png" alt="Agents and Models" width="820">
|
|
114
|
+
|
|
115
|
+
> **Extensions** — Manage global MCP servers, community skills, and built-in automation for headless browsers and macOS desktop (Peekaboo). Add servers via stdio, HTTP, or OAuth 2.1 with Dynamic Client Registration.
|
|
112
116
|
|
|
113
|
-
<img src="docs/promo-dashboard-
|
|
117
|
+
<img src="docs/promo-dashboard-extensions.png" alt="Extensions" width="820">
|
|
114
118
|
|
|
115
|
-
> **System
|
|
119
|
+
> **System & Permissions** — Monitor your working directory alongside real-time CPU, memory, and disk usage, and grant macOS Accessibility / Screen Recording / Disk Access in one place.
|
|
116
120
|
|
|
117
|
-
<img src="docs/promo-dashboard-system.png" alt="System
|
|
121
|
+
<img src="docs/promo-dashboard-system.png" alt="System and Permissions" width="820">
|
|
118
122
|
|
|
119
123
|
</details>
|
|
120
124
|
|
|
@@ -172,7 +176,7 @@ agent CLI versions).
|
|
|
172
176
|
|
|
173
177
|
## How People Are Using It
|
|
174
178
|
|
|
175
|
-
- **Run a Swarm in Parallel** —
|
|
179
|
+
- **Run a Swarm in Parallel** — N panes (or N IM threads), each a different agent on a different repo, all running at once: Claude refactors here, Codex writes tests there, Gemini reviews a third. One cockpit; steer any of them mid-flight.
|
|
176
180
|
- **Self-Hosted Dev Loop** — pikiloom was built using pikiloom. The dev workflow *is* the product: drive the orchestrator from your phone, write code, ship a release, and iterate.
|
|
177
181
|
- **Walk-Away Coding** — Kick off a massive refactoring task, close your laptop, and monitor/steer it from your phone over Telegram. The agent continues running locally, streaming results back to your chat.
|
|
178
182
|
- **Multi-Agent Tag Team** — Let Claude Code draft an initial implementation, switch to Codex for an in-depth review, and finally hand it over to Gemini for a fresh perspective. Same files, same continuous session history.
|