Repository Issue Activity (beta)

#282 OpenAI-compat: AtomicAgent Mode.JSON on factory and AgentConfig, not default tools mode

#275 Typed video input support: VideoURL multimodal class

#232 tooling: improve experience for users of non-Claude-Code AI coding assistants

#233 docs: explore building an MCP server for Atomic Agents documentation

#231 claude-code-plugin: improve dev experience (skills, subagents, structure)

#240 Integration: PaperClaw tool — generate peer-reviewed papers from any Atomic Agents agent

#264 Skills fail to load in GitHub Copilot CLI ≥1.0.65 due to array-shaped `argument-hint`

#262 Make agent memory pluggable via a clear base class (BaseChatHistory) + docs + example

#261 Community tool: DakeraAgentMemory — persistent cross-session memory for atomic agents

#260 Test

#210 Tool provider: Fía Signals crypto intelligence (regime, signals, DeFi yields)

#229 CI governance checks for agent pipelines

#251 Proposal: Agent Identity + Onboarding protocols for interoperable atomic agents

#254 📋 Documentation Enhancement Suggestion

#73 Suggested Framework Enhancements (with example Code)

#241 Doubleword as an inference provider

#230 docs: clarify how tools work in Atomic Agents (no tools=[...] parameter) and add a dedicated guide

#238 atomic-reviewer subagent: two recurring false positives (embeddings + ChatHistory.initialize_turn)

#178 Bug: Instructor hook "parse:error" not emitted

#87 Auto-trim content if too large for context window

#225 chore: update instructor dependency from 1.13.0 to 1.14.5

#221 history.add_message("system", ...) silently dropped mid-conversation when using Gemini backend — breaks ReAct loops

#132 tavily_search.py: max_results is ignored as parameter (fix included)

#193 Enum in agent output fails with is_instance_of validation error

#215 TokenCounter.get_max_tokens() returns output token limit instead of input context window