[NeurIPS 2024] Agent Planning with World Knowledge Model
[ICLR 2025] Benchmarking Agentic Workflow Generation
[ACL 2025] Agentic Knowledgeable Self-awareness
Compound model panel for pi: parallel model calls, then one synthesis response.
[ACL 2025] SynWorld: Virtual Scenario Synthesis for Agentic Action Knowledge Refinement
Least-to-Most decomposition prompt planner organizing complex objectives into sequential subproblems with cumulative state propagation.
Least-to-Most decomposition prompt planner organizing complex objectives into sequential subproblems with cumulative state propagation.
Code for "GraphicWeaver: Benchmarking Agentic Planning for Graphic Design Generation" [ACL ALVR 2026]
Unified benchmark for evaluating architectural design choices in LLM-based single-agent and multi-agent frameworks across orchestration, memory, planning, specialization, and coordination.
Execution-time drift firewall for long-running DeepSeek Harness agents. Real-Harness tests: unsafe stale mutations 12/12 native -> 0/12; valid controls 7/7 both; post-SIGKILL unsafe continuation 2/2 -> 0/2.
Artifact repository for the paper "From Plan to Action: How Well Do Agents Follow the Plan?"
An intuitive, Google Tag Manager-inspired system to embed, govern, and utilize anchor tags for deterministic agent planning and validation in complex codebases
Repo-native operating layer for AI agents: memory, planning, proof, routing, and continuity across sessions.
Trustworthy biomedical tool-using agent with task-conditioned planning, execution-grounded verification, and traceable repair.
Planning-first, edit-later. A reusable skill that brings plan-mode behavior to any AI coding agent — preventing premature edits until you approve.
Compare versioned Markdown plans with JSONL execution traces, review deviations, and record accepted changes or rollback instructions.
Open-source runtime for auditable multi-agent evaluation, starting with professional capital-market investment agents.
Create, parse and validate Markdown scope ledgers that record what a coding task will and will not implement, with explicit rationales.
To associate your repository with the agent-planning topic, visit your repo's landing page and select "manage topics."