Skip to content

docs(agent-testing): correct stale single-turn claims now that conversation: ships - #595

Open
HappyDevs1 wants to merge 3 commits into
mainfrom
docs/agent-conversations-status-update
Open

HappyDevs1 wants to merge 3 commits into
mainfrom
docs/agent-conversations-status-update

Conversation

@HappyDevs1

Copy link
Copy Markdown
Collaborator

Summary

  • PR agent: multi-turn conversation testing (issue #375) #584 shipped conversation: (multi-turn agent flows, issue Agent boundary: support real multi-turn conversations without collapsing prompts #375) with delivery-gated cassette replay for both command: and url: drivers, but several docs still said "a flow is ONE turn, not a conversation" and listed multi-turn conversations under "Not built yet."
  • Updates docs/agent-testing/status-and-scope.md, docs/agent-testing/index.md, docs/adopting.md, and docs/getting-started/record-and-replay.md to describe what conversation: covers today and point at it instead of the old single-turn framing.
  • Preserves the real, still-open gap from plans/012-agent-multiturn-conversations.md: no Anthropic/streaming conversation fixtures yet (OpenAI non-streaming only), and conversation: falls back to the single-shot path when egress containment is also engaged.

Test plan

  • Docs-only change; no code touched.
  • Cross-checked new prose against plans/012-agent-multiturn-conversations.md and crates/flowproof-agent/src/spec.rs for accuracy.

…sation: ships

PR #584 built conversation: (issue #375) with delivery-gated cassette
replay for both drivers, but the docs still said multi-turn wasn't
built. Update status-and-scope.md, index.md, adopting.md, and
record-and-replay.md to describe what conversation: covers today, and
note the real known gap: no Anthropic/streaming fixtures yet, and
conversation: falls back to single-shot when egress containment is
also engaged.
Copilot AI lite review requested due to automatic review settings September 15, 2026 08:44

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

HappyDevs1 and others added 2 commits September 15, 2026 15:39
The dialect/streaming gap this page described is closed: an Anthropic
conversation fixture and a streaming one both exist, so describing them
as outstanding would be the same defect in the other direction. The
coverage table gains the row, including the four red paths.

Egress containment is still a real carve-out and stays.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants