evalsi-harness
No liveness check has reached it yet; it is registered on GitHub and PyPI and has been in the index since 12 Oct 2026. How this is checked
evalsi-harness is an Apache-2.0 Python package for building and evaluating tool-using agents. It provides a tool loop, MCP tool support, sandboxes, a user simulator, and record-and-replay capabilities for developers working with agent evaluations.
Inferred · not functionally tested
Overview
5 featuresPurpose: Building and evaluating tool-using agents with repeatable simulations, sandboxes, and recorded runs.
Inferred · not functionally tested
Audience: AI agent developers and evaluation engineers
Inferred · not functionally tested
Functions: agents, workflow_automation
Inferred · not functionally tested
Interfaces: API: indicated (inferred, not tested) · MCP: indicated (inferred, not tested) · CLI: indicated (inferred, not tested) · Self-hosting: indicated (inferred, not tested)
Recorded constraints: pricing: open_source · license: Apache-2.0 · platforms: CLI · deployment: cli, self_hosted, api_only
Constraint provenance is unknown; confirm requirements with the publisher.
Record sources: pypi.org · github.com. These links do not verify the individual claims.
In the Frameworks & runtimes space, evalsi-harness takes a focused approach. Inferred · not functionally tested: It focuses on building and evaluating tool-using agents with repeatable simulations, sandboxes, and recorded runs. Inferred · not functionally tested: It is built as an open-source project for AI agent developers and evaluation engineers. Basis unknown · not verified: The project is open source (Apache-2.0). Basis unknown · not verified: evalsi-harness is available on the command line and API, and it can be self-hosted.
Behind evalsi-harness is Abhishek-Rnjn, and it first shipped in 2026. The project is developed in the open on GitHub with 249 commits in the last 90 days. Inferred · not functionally tested: Among its 5 catalogued features are tool loop, MCP tools, and sandbox execution. Inferred · not functionally tested: Catalogued interfaces include an MCP server.
Summary written by a language model from the project’s public pages.
Tasks: Inferred · not functionally tested
- Tool loop
- MCP tools
- Sandbox execution
- User simulator
- Record and replay
Topics: Inferred · not functionally tested
Built with & integrations
- Claude Code
- commit 9c22e80a2e4d · since Oct 2026
Trust & compliance
Indexing history
1What PulseGate has recorded for this listing
- Indexed12 Oct · 07:58 UTCevalsi-harness seen via PyPI Bulk EnumeratorSource: PyPI Bulk Enumerator · Open
Frequently asked questions about evalsi-harness
- What is evalsi-harness?
- Inferred · not functionally tested: Evalsi-harness focuses on building and evaluating tool-using agents with repeatable simulations, sandboxes, and recorded runs. It is catalogued under Frameworks & runtimes on PulseGate.
- Who is evalsi-harness for?
- Inferred · not functionally tested: evalsi-harness is an open-source project built for AI agent developers and evaluation engineers.
- Does evalsi-harness have a free plan?
- Basis unknown · not verified: Yes — evalsi-harness is open source under the Apache-2.0 license and free to use.
- What platforms does evalsi-harness run on?
- Basis unknown · not verified: evalsi-harness runs on the command line and API. It can also be self-hosted.
- Is evalsi-harness still active?
- The GitHub repository shows 249 commits in the last 90 days.
- What projects are similar to evalsi-harness?
- Similar projects tracked by PulseGate include harnesslab, skill-eval-harness, and diy-harness.harnesslabskill-eval-harnessdiy-harness
- Who develops evalsi-harness?
- evalsi-harness is developed by Abhishek-Rnjn.
- When did evalsi-harness launch?
- evalsi-harness first shipped in 2026.
Similar projects
Closest matches by what these projects do