PyRank

Agent Reliability Python Packages

Python packages with the GitHub topic agent-reliability. Sorted by relevance, with stars and monthly downloads.
qualixar
superlocalmemory

World's first local-only AI memory to break 74% retrieval and 60% zero-LLM on LoCoMo. No cloud, no APIs, no data leaves your machine. Additionally, mode C (LLM/Cloud) - 87.7% LoCoMo. Research-backed. arXiv: 2603.14588

11K 150 18
kalibr-ai
kalibr

Stop overpaying to run your agents. Kalibr routes every request to lower-cost model and tool paths without degrading performance.

7K 30 3
qualixar
agentassert-abc

Formal behavioral specification and runtime enforcement for autonomous AI agents. Agent Behavioral Contracts (ABC).

966 4 0
Johin2
self-heal-llm

Automatic repair for failing Python code, powered by any LLM.

925 1 5
seekcontext
harness0

Harness engine for AI Agents. From demo to production.

359 0 0
varun369
llama-index-storage-chat-store-superlocalmemory

World's first local-only AI memory to break 74% retrieval and 60% zero-LLM on LoCoMo. No cloud, no APIs, no data leaves your machine. Additionally, mode C (LLM/Cloud) - 87.7% LoCoMo. Research-backed. arXiv: 2603.14588

238 152 19
kalibr-ai
llm-cost-reduction

Stop overpaying to run your agents. Kalibr routes every request to lower-cost model and tool paths without degrading performance.

226 30 3
kalibr-ai
llm-cost-optimizer

Stop overpaying to run your agents. Kalibr routes every request to lower-cost model and tool paths without degrading performance.

217 30 3
yangfei222666-9
self-improving-loop

Rollback-first reliability layer for AI agents: trace failures, apply guarded changes, and rollback on regression.

126 0 0