Metadata-Version: 2.4
Name: brainforge-inference-plugin-history
Version: 0.0.1a1
Summary: Request history step for BrainForge inference pipelines
Home-page: https://github.com/NeonGeckoCom/brainforge-inference-plugin-history
Author: NeonGecko
Author-email: developers@neon.ai
License: AGPL-3.0-only
Classifier: Intended Audience :: Developers
Classifier: Programming Language :: Python :: 3
Description-Content-Type: text/markdown
License-File: LICENSE.md
Requires-Dist: brainforge-inference-utils>=0.0.1a1,~=0.0
Provides-Extra: test
Requires-Dist: pytest~=8.0; extra == "test"
Dynamic: author
Dynamic: author-email
Dynamic: classifier
Dynamic: description
Dynamic: description-content-type
Dynamic: home-page
Dynamic: license
Dynamic: license-file
Dynamic: provides-extra
Dynamic: requires-dist
Dynamic: summary

# BrainForge Inference Plugin: History

Request step that validates and truncates `LLMRequest.history` after earlier
request steps have prepended context. It repairs history rather than skipping:
unknown messages are dropped, prior citation markup is stripped, and
conversational turns are truncated to complete user/llm pairs.

Entry point group: `brainforge.inference.request_steps`

| Step id | Class |
| --- | --- |
| `brainforge.history` | `RequestHistoryStep` |

```yaml
pipeline:
  request:
    - id: brainforge.history
```

Run this step last on the request pipeline. It writes
`transaction.request_step_context["brainforge.history"]`.
