Skip to content

Commit e5cdcb5

Browse files
committed
Sync fxtr plugin 0.1.4
1 parent e04864c commit e5cdcb5

13 files changed

Lines changed: 468 additions & 60 deletions

File tree

‎.claude-plugin/marketplace.json‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -21,7 +21,7 @@
2121
"name": "fxtr",
2222
"source": "./plugins/fxtr",
2323
"description": "Skills for writing, running, and viewing fxtr experiments, and for calling models from them with behaviors",
24-
"version": "0.1.3",
24+
"version": "0.1.4",
2525
"author": {
2626
"name": "TransluceAI"
2727
},

‎plugins/fxtr/.claude-plugin/plugin.json‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -4,5 +4,5 @@
44
"author": {
55
"name": "TransluceAI"
66
},
7-
"version": "0.1.3"
7+
"version": "0.1.4"
88
}

‎plugins/fxtr/skills/behaviors/SKILL.md‎

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
---
22
name: behaviors
3-
description: Call language models from fxtr experiments with the behaviors Python library, covering one-shot requests and judge steps, multi-turn chat rollouts, scripted or custom user and tool policies, retries and failures, streaming, checkpoints, and storing conversations. Use whenever a fxtr experiment calls a model or holds a conversation, or when extending behaviors' model adapters or policies.
3+
description: Call language models from fxtr experiments with the behaviors Python library, covering one-shot requests and judge steps, multi-turn chat rollouts, scripted, model-simulated, or custom users and tool policies, retries and failures, streaming, checkpoints, and storing conversations. Use whenever a fxtr experiment calls a model or holds a conversation, or when extending behaviors' model adapters or policies.
44
---
55

66
# behaviors
@@ -18,11 +18,11 @@ The pages below are its documentation. Read the ones a task needs before writing
1818
| ------------------------------------------------------------------------------------------------------------------------------ | ---------------------------------------------------------------------------------- |
1919
| Add behaviors to a project; hold a conversation, judge outputs, and summarize verdicts in fxtr steps; test without credentials | [Calling language models](references/guides/language-models.mdx) |
2020
| Pick a provider, an adapter, and an exact model ID | [Choosing models](references/guides/choosing-models.mdx) |
21-
| Script users, offer tools, or write a custom conversation flow | [Policies and tools](references/guides/policies-and-tools.mdx) |
21+
| Script users or simulate them with models (a base model, steered by a pilot), offer tools, or write a custom conversation flow | [Policies and tools](references/guides/policies-and-tools.mdx) |
2222
| Rollout statuses and `max_turns`, failures, retries, streaming | [Rollout outcomes and retries](references/guides/rollout-outcomes.mdx) |
2323
| Messages and content, stored records, transcripts, conversations in the viewer, checkpoints | [Conversation records and checkpoints](references/guides/conversation-records.mdx) |
2424
| Run Docent readings, or convert records for Docent | [Docent readings](references/guides/docent-readings.mdx) |
25-
| Choose a layer; imports, adapters, sampling settings, requests and responses | [API overview](references/reference/api.mdx) |
25+
| Choose a layer; imports, adapters and their options, model providers, sampling and reasoning settings, requests and responses | [API overview](references/reference/api.mdx) |
2626
| Steps, workflows, arrays, launching jobs, and the step cache | the adjacent [fxtr skill](../fxtr/SKILL.md) |
2727

2828
## Best practices

‎plugins/fxtr/skills/behaviors/provenance.json‎

Lines changed: 4 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -2,10 +2,10 @@
22
"skill": "behaviors",
33
"documentation": "https://docs.transluce.ai/behaviors",
44
"packages": {
5-
"fxtr": "0.0.1a4",
6-
"behaviors": "0.0.1a3"
5+
"fxtr": "0.1.0a0",
6+
"behaviors": "0.0.1a4"
77
},
8-
"revision": "c3782b939f40b8fbffa00c43c984c0ff83087cb6",
8+
"revision": "052f907cc35979930fda832a71aec0c2194e7e48",
99
"modified": false,
10-
"content_sha256": "9ca8c278e92e77f29f281313b17717930e6ecc67e13027fedde3152fc37ea5b9"
10+
"content_sha256": "a9a5b26e4b86c7b765f0d3c7ce1ce7bea062364e7112c5bec2ad86aed6ca2f1d"
1111
}

‎plugins/fxtr/skills/behaviors/references/INDEX.md‎

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -14,6 +14,6 @@ in the order of the site's navigation.
1414
- [Conversation records and checkpoints](guides/conversation-records.mdx): What a rollout records, how conversations are stored as fxtr entities and flattened into transcripts, and how to checkpoint a session.
1515
- [Docent readings](guides/docent-readings.mdx): Convert conversations to Docent's data models, and run a Docent reading against a transcript or agent run with a model you choose.
1616
- [Calling language models](guides/language-models.mdx): Use the behaviors library to hold conversations and judge outputs in fxtr steps.
17-
- [Policies and tools](guides/policies-and-tools.mdx): Decide what the model sees each turn: scripted users, tool policies, their combinations, and custom conversation flows.
17+
- [Policies and tools](guides/policies-and-tools.mdx): Decide what the model sees each turn: scripted users, users simulated by a base model and steered by a pilot, tool policies, their combinations, and custom conversation flows.
1818
- [Rollout outcomes and retries](guides/rollout-outcomes.mdx): How a rollout ends, how model call failures are retried, recorded, or raised, and how to watch a rollout as it runs.
19-
- [API overview](reference/api.mdx): The layers of behaviors, where each is imported from, the provider adapters, and the request, response, and sampling types.
19+
- [API overview](reference/api.mdx): The layers of behaviors, where each is imported from, the provider adapters and model providers, and the request, response, and sampling types.

‎plugins/fxtr/skills/behaviors/references/guides/conversation-records.mdx‎

Lines changed: 11 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -132,6 +132,14 @@ A new project's renderer package doesn't depend on it yet: add
132132
[The renderer package](../../../fxtr/references/reference/renderers.mdx#the-renderer-package)). The package also exports
133133
`ChatRolloutView`, for a rollout inside a view of your own, and `ComparedRollouts`.
134134

135+
A rollout's or a transcript's conversation view reads a `selector` param, `{ message: n }`, and
136+
marks and scrolls to that message: the `n`th of the messages it draws, counting a transcript's
137+
messages in order, and a rollout's turn by turn, each turn's context messages and then its model's
138+
(not its initialization). A view of your own links to it with that selector, read from an entity
139+
that stores one (see [Params of an entity's
140+
views](../../../fxtr/references/reference/renderers.mdx#params-of-an-entitys-views)); `MessageSelector` and
141+
`isMessageSelector` are exported for it. This selector is provisional.
142+
135143
## Checkpointing sessions
136144

137145
A serializable session can save its state and be restored from it. `await session.get_state()`
@@ -140,7 +148,9 @@ returns the state as `DynamicData`, tagged with the implementation's own state t
140148
for a state that isn't its own. To keep a state as JSON, write
141149
`json.dumps(thaw_fxtr_data(state.data))` (`thaw_fxtr_data` is in `fxtr.core.entity_data`), and
142150
read it back with `DynamicData.from_data(json.loads(text))`.
143-
The sessions of `chat_session_model_from_api` and of `ScriptedUserSimulator` support this, and the
151+
The sessions of `chat_session_model_from_api`, `ScriptedUserSimulator`, and the
152+
[simulated users](policies-and-tools.mdx#checkpointing-a-simulated-user)
153+
(`BaseModelUserSimulator` and `PilotSteeredUserSimulator`) support this, and the
144154
`serializable_context_policy_from_*` combinators build checkpointable policies when every component
145155
they combine is checkpointable. Generator-based policies can't be checkpointed, because a
146156
suspended generator can't be saved.

‎plugins/fxtr/skills/behaviors/references/guides/language-models.mdx‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -7,7 +7,7 @@ description: "Use the behaviors library to hold conversations and judge outputs
77

88
* model adapters for Anthropic, OpenAI, and OpenAI-compatible services;
99
* conversation sessions and **context policies**, which decide what the model sees next:
10-
scripted user messages, tool results, or custom logic;
10+
scripted or model-simulated user messages, tool results, or custom logic;
1111
* retries for rate limits and server errors;
1212
* conversation records that are fxtr entities, so they're stored, linked, and shown in the
1313
viewer.

0 commit comments

Comments
 (0)