v1.0.0

First stable release. Since 0.6: the schema follows the A2A v1 specification; message/stream and tasks/resubscribe stream over SSE through an event bus, and a task's stream ends when its worker is done; the Pydantic AI bridge lives here as fasta2a.pydantic_ai and streams the answer as artifact chunks; agents declare protocol extensions and negotiate them per request through the A2A-Extensions header.

What's Changed

• fix(schema): update authentication field in PushNotificationAuthenticationInfo by @Kludex in #32
• Add docs_url parameter by @awickham10 in #35
• Update AgentCard and related types to A2A v1 spec by @Kludex in #46
• Update schema types for A2A v1 spec (Phase 2) by @Kludex in #47
• Add StreamResponse wrapper type for A2A v1 spec by @Kludex in #48
• Update push notification types and add ListTasks for A2A v1 spec by @Kludex in #49
• Route all A2A methods and return proper JSON-RPC error responses by @Kludex in #50
• Implement SSE streaming for message/stream and tasks/resubscribe by @Kludex in #51
• Add pydantic_ai bridge submodule (port from pydantic-ai _a2a.py) by @dsfaccini in #55
• Fix docs API paths for mounted apps by @ChenXiaoTemp in #59
• feat: negotiate A2A extensions by @echarles in #61
• feat: end a task's stream when its worker is done, and stream the bridge's answer by @echarles in #62

New Contributors

• @awickham10 made their first contribution in #35
• @ChenXiaoTemp made their first contribution in #59
• @echarles made their first contribution in #61

Full Changelog: v0.6.0...v1.0.0

v0.6.1

Maintenance release on the pre-A2A-v1-spec line.

What's new

• Adds the fasta2a.pydantic_ai bridge submodule (agent_to_a2a, AgentWorker, worker_lifespan) — a port of pydantic_ai._a2a from the Pydantic AI repo, so users have a migration target for Agent.to_a2a() before its deprecation in Pydantic AI v2.
• New [pydantic-ai] optional extra: pip install 'fasta2a[pydantic-ai]' (requires Python 3.10+).

from pydantic_ai import Agent
from fasta2a.pydantic_ai import agent_to_a2a

agent = Agent('openai:gpt-4.1')
app = agent_to_a2a(agent)

Companion change on Pydantic AI side: pydantic/pydantic-ai#5426

Notes

This release ships from release/0.6, branched from the last commit before the A2A v1 spec rewrite started. The full v1 spec rewrite remains on main and will land in a later release.

Full diff: v0.6.0...v0.6.1

v0.6.0 (2025-10-07)

What's Changed

• Add StreamMessageResponse to schema.py by @izzymsft in #17
• Adding code documentation to schema.py by @izzymsft in #23
• Support protocolVersion 0.3.0 by @noklam in #31

New Contributors

• @izzymsft made their first contribution in #17
• @noklam made their first contribution in #31

Full Changelog: v0.5.0...v0.6.0

v0.5.0 (2025-07-10)

This is the first release in this repository! 🎉

What's Changed

• Add CI by @Kludex in #4
• Add .gitignore by @Kludex in #5
• Run CI on main by @Kludex in #6
• Add linting rules by @Kludex in #7
• Add CLAUDE.md by @Kludex in #8
• Add /docs endpoint by @Kludex in #9
• Add documentation by @Kludex in #10
• Drop hatchling versioning inference from publish by @Kludex in #11

New Contributors

• @Kludex made their first contribution in #4

Full Changelog: https://github.com/pydantic/fasta2a/commits/v0.5.0