tagteam 0.2.0 → 0.4.1

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 CHANGED
@@ -53,14 +53,49 @@ You can also set the default pair in your config:
53
53
  agents = ["claude", "gemini"]
54
54
  ```
55
55
 
56
+ ### Addressing agents inline
57
+
58
+ Prefix your prompt with an agent name to send it to just that agent:
59
+
60
+ ```
61
+ gemini what do you think about this approach?
62
+ claude summarize the discussion so far
63
+ ```
64
+
65
+ Name two agents to override the active pair for that prompt (separators: `and`, `/`, `&`, `,`):
66
+
67
+ ```
68
+ gemini and codex review this function
69
+ gemini/claude compare your approaches
70
+ gemini & codex review this function
71
+ ```
72
+
73
+ Add `discuss` to start a multi-round discussion — either word order works:
74
+
75
+ ```
76
+ discuss codex and gemini what's the best caching strategy
77
+ codex and gemini discuss what's the best caching strategy
78
+ ```
79
+
80
+ This works for any agent, not just your active pair. Addressed agents receive the full conversation history for context.
81
+
56
82
  ### Discussion mode
57
83
 
58
- Have your agents discuss a topic in rounds until they reach consensus:
84
+ Have your agents discuss a topic in structured rounds with research-backed prompting:
59
85
 
60
86
  ```bash
61
87
  tagteam discuss "what's the best approach for caching in this app"
62
88
  ```
63
89
 
90
+ Discussions use a structured debate protocol informed by multi-agent debate research (Du et al. ICML 2024, ReConcile ACL 2024, CONSENSAGENT ACL Findings 2025, MAD EMNLP 2024):
91
+
92
+ - **Role differentiation** — each agent has a distinct role (Builder, Verifier, or Strategist) with specific focus areas, and sees an anonymized description of their peer's role
93
+ - **Anti-sycophancy rules** — agents must justify position changes with specific arguments and introduce novel content each round
94
+ - **Structured arguments** — debate responses follow Toulmin structure (claim, evidence, reasoning, caveats) with confidence and position-change markers
95
+ - **Steelman injection** — in round 2, agents are prompted to steelman the opposing view before responding
96
+ - **Smart termination** — discussions end early on mutual consensus, stale debates (no new arguments for 2 rounds), or cyclic position-swapping, with a clear status message explaining why
97
+ - **Context summarization** — rounds 3+ receive a compressed summary of earlier rounds plus full text of the latest exchange, keeping prompts focused
98
+
64
99
  ### Session management
65
100
 
66
101
  ```bash
@@ -70,6 +105,12 @@ tagteam continue
70
105
  # List recent sessions
71
106
  tagteam history
72
107
 
108
+ # Delete a specific session
109
+ tagteam history rm <id>
110
+
111
+ # Delete all sessions
112
+ tagteam history clear
113
+
73
114
  # Resume a specific session by ID
74
115
  tagteam resume <id>
75
116
 
@@ -82,7 +123,7 @@ tagteam show <id> --markdown
82
123
 
83
124
  ### Configuration
84
125
 
85
- Configuration is stored in `~/.tagteam/config.toml` (or `%APPDATA%\tagteam\config.toml` on Windows).
126
+ Configuration is stored in `~/.tagteam/config.toml` (or `%APPDATA%\tagteam\config.toml` on Windows). Session data is stored in `~/.tagteam/tagteam.db`.
86
127
 
87
128
  ```bash
88
129
  # Interactive config editor
@@ -96,9 +137,19 @@ tagteam config set agents claude,gemini
96
137
  tagteam config set claude.model sonnet
97
138
  tagteam config set codex.model gpt-5.3-codex
98
139
  tagteam config set gemini.model gemini-2.5-pro
99
- tagteam config set discussion.max_rounds 10
140
+ tagteam config set discussion.max_rounds 5
100
141
  ```
101
142
 
143
+ Default values:
144
+
145
+ | Key | Default |
146
+ |-----|---------|
147
+ | `agents` | `claude, codex` |
148
+ | `claude.model` | `sonnet` |
149
+ | `codex.model` | `gpt-5.3-codex` |
150
+ | `gemini.model` | `gemini-2.5-pro` |
151
+ | `discussion.max_rounds` | `5` |
152
+
102
153
  ### CLI options
103
154
 
104
155
  ```bash