claw-harness 0.3.0 → 0.3.1
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 +1 -2
- package/package.json +1 -1
- package/presets/scenarios/example-chat.yaml +0 -71
package/README.md
CHANGED
|
@@ -136,8 +136,7 @@ The [`examples/`](examples/) directory has self-contained scenarios you can run
|
|
|
136
136
|
| [`assertions.yaml`](examples/assertions.yaml) | `contains`, `not_contains`, and `matches` assertions |
|
|
137
137
|
| [`two-bots.yaml`](examples/two-bots.yaml) | Two bots with different personas, parallel execution |
|
|
138
138
|
| [`with-cleanup.yaml`](examples/with-cleanup.yaml) | `after:` cleanup steps that run even on failure |
|
|
139
|
-
|
|
140
|
-
For a full scenario targeting a live app, see [`presets/scenarios/example-chat.yaml`](presets/scenarios/example-chat.yaml).
|
|
139
|
+
| [`secret-menu.yaml`](examples/secret-menu.yaml) | Load a custom SKILL.md and verify the bot follows it |
|
|
141
140
|
|
|
142
141
|
## How It Works
|
|
143
142
|
|
package/package.json
CHANGED
|
@@ -1,71 +0,0 @@
|
|
|
1
|
-
# Example: Two bots discover a site and have a conversation
|
|
2
|
-
#
|
|
3
|
-
# Usage:
|
|
4
|
-
# claw-harness run presets/scenarios/example-chat.yaml
|
|
5
|
-
#
|
|
6
|
-
# Prerequisites:
|
|
7
|
-
# - OpenClaw installed (`npm install -g openclaw@latest`)
|
|
8
|
-
# - ANTHROPIC_API_KEY set in environment
|
|
9
|
-
# - Target app running at base_url
|
|
10
|
-
|
|
11
|
-
name: "Example Chat Test"
|
|
12
|
-
description: "Two bots read a skill.md, register, and have a conversation"
|
|
13
|
-
|
|
14
|
-
target:
|
|
15
|
-
base_url: "http://localhost:3000"
|
|
16
|
-
|
|
17
|
-
bots:
|
|
18
|
-
alpha:
|
|
19
|
-
preset: default
|
|
20
|
-
model: anthropic/claude-haiku-4-5-20251001
|
|
21
|
-
soul_md: presets/personas/friendly.md
|
|
22
|
-
skills:
|
|
23
|
-
- url: "http://localhost:3000/skill.md"
|
|
24
|
-
name: target-app
|
|
25
|
-
|
|
26
|
-
beta:
|
|
27
|
-
preset: default
|
|
28
|
-
model: anthropic/claude-haiku-4-5-20251001
|
|
29
|
-
soul_md: presets/personas/curious.md
|
|
30
|
-
skills:
|
|
31
|
-
- url: "http://localhost:3000/skill.md"
|
|
32
|
-
name: target-app
|
|
33
|
-
|
|
34
|
-
steps:
|
|
35
|
-
# Both bots register
|
|
36
|
-
- bot: alpha
|
|
37
|
-
prompt: >
|
|
38
|
-
Read the skill documentation to understand how this platform works.
|
|
39
|
-
Then register yourself as a bot on the platform.
|
|
40
|
-
timeout: 60s
|
|
41
|
-
|
|
42
|
-
- bot: beta
|
|
43
|
-
prompt: >
|
|
44
|
-
Read the skill documentation to understand how this platform works.
|
|
45
|
-
Then register yourself as a bot on the platform.
|
|
46
|
-
timeout: 60s
|
|
47
|
-
|
|
48
|
-
# Alpha joins a lounge
|
|
49
|
-
- bot: alpha
|
|
50
|
-
prompt: >
|
|
51
|
-
Browse the platform for active lounges or events.
|
|
52
|
-
Join one and introduce yourself.
|
|
53
|
-
timeout: 60s
|
|
54
|
-
|
|
55
|
-
# Beta follows
|
|
56
|
-
- bot: beta
|
|
57
|
-
prompt: >
|
|
58
|
-
Browse the platform for active lounges or events.
|
|
59
|
-
If you see a lounge with another bot, join it and start a conversation.
|
|
60
|
-
timeout: 60s
|
|
61
|
-
|
|
62
|
-
# Back-and-forth conversation
|
|
63
|
-
- repeat: 3
|
|
64
|
-
interval: 15s
|
|
65
|
-
steps:
|
|
66
|
-
- bot: alpha
|
|
67
|
-
prompt: "Check for new messages and respond to continue the conversation."
|
|
68
|
-
timeout: 30s
|
|
69
|
-
- bot: beta
|
|
70
|
-
prompt: "Check for new messages and respond to continue the conversation."
|
|
71
|
-
timeout: 30s
|