PR #11738: fix(acp): accept wrapped prompt payloads

• Repository: NousResearch/hermes-agent
• Author: Lucenx9
• State: open | merged: False
• Link: https://github.com/NousResearch/hermes-agent/pull/11738

Description (problem / solution / changelog)

What does this PR do?

Fixes #11732: some ACP clients can call the adapter with a request-like wrapper object instead of passing the raw prompt block list directly. The current ACP adapter assumes prompt is always iterable, so _extract_text() crashes with TypeError: 'types.SimpleNamespace' object is not iterable before Hermes ever reaches the model.

This PR keeps the fix narrow to the ACP boundary:

• accept prompt payloads passed either as the normal block list or as a wrapper object exposing .prompt
• keep the existing text extraction behavior once the block list is normalized
• add a regression test that exercises agent.prompt(prompt=SimpleNamespace(prompt=[...]))

Related Issue

Fixes #11732

Type of Change

• 🐛 Bug fix (non-breaking change that fixes an issue)
• ✨ New feature (non-breaking change that adds functionality)
• 🔒 Security fix
• 📝 Documentation update
• ✅ Tests (adding or improving test coverage)
• ♻️ Refactor (no behavior change)
• 🎯 New skill (bundled or hub)

Changes Made

• acp_adapter/server.py: normalize ACP prompt payloads that arrive as wrapper objects with a .prompt attribute before iterating blocks in _extract_text()
• tests/acp/test_server.py: add a regression test proving wrapped prompt payloads no longer crash and still call run_conversation() with the extracted text

How to Test

1. Reproduce the old failing shape by calling the ACP adapter with a wrapped payload, e.g. prompt=SimpleNamespace(prompt=[TextContentBlock(type="text", text="hello")])
2. Confirm the ACP adapter no longer raises TypeError and instead forwards hello into run_conversation()
3. Run: pytest -o addopts='' tests/acp/test_server.py -q
4. Run: pytest -o addopts='' tests/acp -q

Checklist

Code

• I've read the Contributing Guide
• My commit messages follow Conventional Commits (fix(scope):, feat(scope):, etc.)
• I searched for existing PRs to make sure this isn't a duplicate
• My PR contains only changes related to this fix/feature (no unrelated commits)
• I've run pytest tests/ -q and all tests pass
• I've added tests for my changes (required for bug fixes, strongly encouraged for features)
• I've tested on my platform: Linux (local dev environment)

Documentation & Housekeeping

• I've updated relevant documentation (README, docs/, docstrings) — or N/A
• I've updated cli-config.yaml.example if I added/changed config keys — or N/A
• I've updated CONTRIBUTING.md or AGENTS.md if I changed architecture or workflows — or N/A
• I've considered cross-platform impact (Windows, macOS) per the compatibility guide — or N/A
• I've updated tool descriptions/schemas if I changed tool behavior — or N/A

Screenshots / Logs

Validation run:

• pytest -o addopts='' tests/acp/test_server.py -q → 42 passed
• pytest -o addopts='' tests/acp -q → 128 passed

Changed files

• acp_adapter/server.py (modified, +13/-1)
• tests/acp/test_server.py (modified, +28/-0)

PR #11780: fix(streaming): handle final responses from non-streaming adapters

• Repository: NousResearch/hermes-agent
• Author: LeonSGP43
• State: open | merged: False
• Link: https://github.com/NousResearch/hermes-agent/pull/11780

Description (problem / solution / changelog)

Summary

• detect adapters that accept \ but return a final response object
• disable streaming for the rest of the session instead of iterating a non-stream response
• cover the ACP Copilot regression with a focused streaming fallback test

Testing

• python3 -m pytest -o addopts='' tests/run_agent/test_streaming.py -k 'text_only_response or response_object_disables_streaming_and_returns_final_response'

Closes #11732

Changed files

• run_agent.py (modified, +29/-0)
• tests/run_agent/test_streaming.py (modified, +47/-0)