Repository Issue Activity (beta)

#323 readurls approach fetches attacker-named URLs with no SSRF guard (metadata/internal reachable on hosted proxy)

#327 Using AnyLLM as a provider — cost-efficient Chinese models + OptiLLM techniques (bon/moa/mcts on DeepSeek, Qwen, Kimi)

#325 proposal: Integration or tracer hook for Agent-Devtools / observability

#275 Use linters in your project please

#111 add file store for memory plugin

#304 Bug: Race condition — concurrent requests corrupt MCTS parameters via shared global state

#312 none approach drops tool_calls on streaming requests

#308 [Feature] Anthropic-compatible API

#249 Add a compact plugin that does auto compact on context like claude code

#306 Where to put local LLM models for Local inference server ?

#303 Security: RCE via prompt injection in z3_solver — exec() on LLM-generated code

#302 Security: Unauthenticated RCE via executecode plugin — unsandboxed Python execution

#299 Persistent Agent Memory & Cross-Session Decision Reuse in optillm

#292 json response error

#293 Proposal: Implement RLM

#291 rate limit

#281 Failed to setup Chrome driver

#287 [Critical Security] Remote Code Execution (RCE) Vulnerability Detected

#286 zai llm

#278 optillm/plansearch.py bug

#276 missing google library

#272 LM-Studio as endpoint, error 400 Cannot read properties of null ( reading 'type' )

#269 Docker `latest` tag not updating to v0.3.5

#264 SPL plugin not working

#262 Usage of `max_tokens` is ambiguous, deprecated and causing errors.