v0.6.2

Remove the legacy License classifier now that PEP 639 license_expression is set (setuptools>=78 forbids both together).

v0.6.1

Docs-only release: strips the archived TRAIL/Who&When benchmark section and the "32 core detectors" count claims from README.md. No code changes.

v0.6.0

Added

• pisama.auto — consolidated in-package original of the standalone pisama-auto distribution (zero-code auto-instrumentation for the Anthropic/OpenAI SDKs, tracing to either the Pisama platform or a self-hosted OTLP collector). New auto extra: pip install "pisama[auto]".
• pisama.agents — consolidated in-package original of the standalone pisama-agent-sdk distribution (real-time hooks, tools, and a pisama-openhands-monitor console script for Claude Agent SDK / Harbor integrations). New agents and telemetry extras: pip install "pisama[agents]".
• pisama.agents.__version__, aliased to pisama.__version__ since this code no longer releases independently.

Both submodules keep publishing as standalone distributions (pisama-auto, pisama-agent-sdk) for this release; those packages become thin shims re-exporting this module in a later release. A bare pip install pisama pulls in no new required dependencies — pisama.auto and pisama.agents stay behind their respective extras.

Also folds in the unreleased 0.5.7 changes (TRAIL precision-boundary disclosure in the README).

See #14.

v0.5.6

Ships two fixes that were merged to main but could not reach users until a release carried them.

Fixed: pisama[mcp] install was broken. The extra declared mcp>=1.0.0 with no upper bound, so a fresh install resolved mcp 2.0.0 and pisama mcp-server crashed on startup:

File ".../pisama/mcp/server.py", line 138, in create_local_server
  @server.list_tools()
AttributeError: 'Server' object has no attribute 'list_tools'

pisama/mcp/server.py uses the mcp 1.x decorator API, which 2.0.0 removed. Now pinned to mcp>=1.0.0,<2. The bound should be lifted once the server is ported to the 2.x API.

Fixed: incorrect README claims. PyPI renders the README from the published artifact, so the project page carried three errors:

• Four documented detectors do not exist. delegation, grounding, retrieval_quality and compaction_quality raise UnknownDetectorError. Replaced with four that do exist and were undocumented: propagation, citation, routing, mcp_protocol.
• "Framework-specific detectors for n8n, LangGraph, Dify, and OpenClaw" was false. There are 32 core detectors, gated per platform; none is unique to a framework.
• The TRAIL benchmark table lacked its claim boundary. The archived run is in-distribution and not held out (144 of 148 traces appeared in calibration material), and the public archive does not contain the prediction-level data needed to recompute joint accuracy. The table now reports only what the archive supports, with the caveat inline and macro-F1 0.7535 / micro-F1 0.7463 named as the reproducible figures.

No API changes.

pisama 0.5.5

What's Changed

• Harden Python package publishing by @tn-pisama in #8
• Prepare pisama 0.5.5 release by @tn-pisama in #9

Full Changelog: v0.5.4...v0.5.5

pisama 0.5.3

What's Changed

• Require cross-version core timestamp parsing by @tn-pisama in #4

New Contributors

• @tn-pisama made their first contribution in #4

Full Changelog: https://github.com/Pisama-AI/pisama-python/commits/v0.5.3