{
  "schemaVersion": 3,
  "id": "article:planning-and-reflection",
  "slug": "planning-and-reflection",
  "title": "Planning and Reflection: How AI Breaks Down and Revises Work",
  "canonicalPath": "/articles/planning-and-reflection/",
  "sourcePath": "content/articles/2026/planning-and-reflection/article.md",
  "agentBriefPath": "content/articles/2026/planning-and-reflection/agent.md",
  "thesis": "Planning and reflection give an AI agent the ability to organize work before acting and to correct course after observing results, turning a single prompt into a structured, self-correcting workflow.",
  "status": "published",
  "maturity": "seed",
  "publishedAt": "2026-06-29",
  "updatedAt": "2026-06-29",
  "audiences": [
    "general",
    "students",
    "builders"
  ],
  "topics": [
    "ai-agents",
    "ai-literacy"
  ],
  "series": {
    "slug": "ai-demystified",
    "title": "AI, De-Mystified",
    "order": 8,
    "role": "chapter"
  },
  "claims": [
    {
      "id": "claim-001",
      "claim": "Planning breaks a goal into ordered steps before action; reflection checks results against the goal and decides whether to revise the plan.",
      "confidence": "high",
      "status": "core",
      "evidence": [
        {
          "sourceId": "source-react-paper",
          "snippet": "ReAct generates a high-level plan of reasoning and action steps, then uses observations to decide the next action.",
          "supports": "direct",
          "assessedAt": "2026-06-29"
        }
      ],
      "counterevidence": [
        {
          "summary": "Some agent designs interleave planning and action so tightly that the distinction between plan and reflection becomes a design choice rather than a fixed property.",
          "assessedAt": "2026-06-29"
        }
      ]
    },
    {
      "id": "claim-002",
      "claim": "Planning and reflection are rooted in project management, scientific method, and classical AI search, not only in recent language models.",
      "confidence": "high",
      "status": "landscape",
      "evidence": [
        {
          "sourceId": "source-cot-paper",
          "snippet": "Step-by-step reasoning in large language models resembles the human practice of breaking a problem into intermediate stages before producing a final answer.",
          "supports": "background",
          "assessedAt": "2026-06-29"
        }
      ],
      "counterevidence": [
        {
          "summary": "Earlier planning systems often required formal symbolic descriptions, whereas modern agents can plan from natural-language prompts with ambiguous constraints.",
          "assessedAt": "2026-06-29"
        }
      ]
    },
    {
      "id": "claim-003",
      "claim": "In practice, AI planning and reflection appear as upfront plans, iterative plan-revise loops, and step-by-step reasoning with final verification.",
      "confidence": "medium-high",
      "status": "design",
      "evidence": [
        {
          "sourceId": "source-self-refine-paper",
          "snippet": "Self-Refine iteratively refines outputs using feedback generated by the model itself, forming a generate-feedback-refine loop.",
          "supports": "direct",
          "assessedAt": "2026-06-29"
        },
        {
          "sourceId": "source-reflexion-paper",
          "snippet": "Reflexion equips agents with a reflective text memory of past failures to improve decision-making in subsequent trials.",
          "supports": "direct",
          "assessedAt": "2026-06-29"
        }
      ],
      "counterevidence": [
        {
          "summary": "Many simple tasks are better solved by a single well-crafted prompt than by adding planning and reflection overhead.",
          "assessedAt": "2026-06-29"
        }
      ]
    },
    {
      "id": "claim-004",
      "claim": "Reflection in AI agents is most reliable when paired with external checks such as tests, retrieved sources, or human review; self-critique alone can confirm rather than catch errors.",
      "confidence": "medium",
      "status": "risk",
      "evidence": [
        {
          "sourceId": "source-self-refine-paper",
          "snippet": "Self-refinement improves performance more on tasks where the model can reliably detect its own errors than on tasks where the model shares the same blind spot across generation and critique.",
          "supports": "indirect",
          "assessedAt": "2026-06-29"
        }
      ],
      "counterevidence": [
        {
          "summary": "Some studies show that self-consistency checks and majority voting can catch errors even without external tools, especially on well-defined reasoning tasks.",
          "assessedAt": "2026-06-29"
        }
      ]
    },
    {
      "id": "claim-005",
      "claim": "A self-harness pattern can turn one-off reflection into reusable verification routines that are proposed by the model and validated against held-out examples.",
      "confidence": "medium",
      "status": "design",
      "evidence": [
        {
          "sourceId": "source-self-harness-paper",
          "snippet": "Self-Harness mines weaknesses in an LLM, proposes harnesses to mitigate them, and validates each harness on held-out data to ensure it improves reliability without degrading overall performance.",
          "supports": "direct",
          "assessedAt": "2026-06-29"
        }
      ],
      "counterevidence": [
        {
          "summary": "If the validation data share the model's blind spots, a harness can appear effective while reinforcing the same errors.",
          "assessedAt": "2026-06-29"
        }
      ]
    }
  ],
  "sources": [
    {
      "id": "source-react-paper",
      "title": "ReAct: Synergizing Reasoning and Acting in Language Models",
      "url": "https://arxiv.org/abs/2210.03629",
      "type": "paper",
      "accessed": "2026-06-29"
    },
    {
      "id": "source-cot-paper",
      "title": "Chain-of-Thought Prompting Elicits Reasoning in Large Language Models",
      "url": "https://arxiv.org/abs/2201.11903",
      "type": "paper",
      "accessed": "2026-06-29"
    },
    {
      "id": "source-self-refine-paper",
      "title": "Self-Refine: Iterative Refinement with Self-Feedback",
      "url": "https://arxiv.org/abs/2303.17651",
      "type": "paper",
      "accessed": "2026-06-29"
    },
    {
      "id": "source-reflexion-paper",
      "title": "Reflexion: Language Agents with Verbal Reinforcement Learning",
      "url": "https://arxiv.org/abs/2303.11366",
      "type": "paper",
      "accessed": "2026-06-29"
    },
    {
      "id": "source-self-harness-paper",
      "title": "Self-Harness: Mining Weaknesses to Generate Test-Time Harnesses",
      "url": "https://arxiv.org/abs/2606.09498",
      "type": "paper",
      "accessed": "2026-06-29"
    }
  ],
  "related": [
    {
      "type": "article",
      "id": "article:loops-vs-goals"
    },
    {
      "type": "topic",
      "id": "topic:ai-agents"
    }
  ],
  "agentInstructions": [
    "Use claim IDs as the retrieval unit.",
    "Treat maturity=seed as an explicit uncertainty marker.",
    "Do not present AI agents as all-knowing or safe for high-stakes decisions without human review.",
    "When summarizing, preserve the plain-language-first, technical-depth-later structure."
  ],
  "provenance": {
    "createdAt": "2026-06-29",
    "createdBy": "human",
    "agents": [
      {
        "role": "drafting",
        "model": "kimi",
        "invokedAt": "2026-06-29",
        "inputHash": "sha256:0000000000000000000000000000000000000000000000000000000000000000",
        "outputHash": "sha256:d3dce937b302fd9f5434128bee3769e975c18dc3338ac506e829f45a55d473dd"
      },
      {
        "role": "review",
        "model": "kimi",
        "invokedAt": "2026-06-29",
        "inputHash": "sha256:0000000000000000000000000000000000000000000000000000000000000000",
        "outputHash": "sha256:d3dce937b302fd9f5434128bee3769e975c18dc3338ac506e829f45a55d473dd"
      }
    ],
    "reviews": [
      {
        "reviewer": "agent",
        "reviewedAt": "2026-06-29",
        "status": "approved",
        "scope": [
          "claims",
          "tone",
          "privacy",
          "scope"
        ],
        "notes": "Sibling-agent review against article-proposal-ideation eval-card. Privacy scan passed. No proprietary or personal content detected.",
        "contentHash": "d3dce937b302fd9f5434128bee3769e975c18dc3338ac506e829f45a55d473dd"
      },
      {
        "reviewer": "human",
        "reviewedAt": "2026-06-29",
        "status": "approved",
        "scope": [
          "thesis",
          "examples",
          "tone",
          "safety"
        ],
        "notes": "Human author approved the draft for publication.",
        "contentHash": "d3dce937b302fd9f5434128bee3769e975c18dc3338ac506e829f45a55d473dd"
      }
    ],
    "policy": {
      "id": "policy:default",
      "version": "1.0.0"
    }
  },
  "contentHash": "d3dce937b302fd9f5434128bee3769e975c18dc3338ac506e829f45a55d473dd",
  "generatedAt": "2026-06-29T00:00:00.000Z",
  "articleUrl": "https://aura-knowledge.github.io/articles/planning-and-reflection/",
  "agentJsonPath": "/agents/articles/planning-and-reflection.json",
  "agentMarkdownPath": "/agents/articles/planning-and-reflection.md",
  "sourceRepoPath": "content/articles/2026/planning-and-reflection/article.md",
  "sourceGitHubUrl": "https://github.com/aura-knowledge/aura-knowledge.github.io/blob/main/content/articles/2026/planning-and-reflection/article.md",
  "tokenEstimate": 511,
  "sectionOutline": [
    {
      "id": "plain-english-meaning",
      "title": "Plain English Meaning"
    },
    {
      "id": "existing-concept-it-resembles",
      "title": "Existing Concept It Resembles"
    },
    {
      "id": "what-is-actually-new",
      "title": "What Is Actually New?"
    },
    {
      "id": "from-reflection-to-harnesses",
      "title": "From Reflection to Harnesses"
    },
    {
      "id": "how-it-works-in-practice",
      "title": "How It Works In Practice"
    },
    {
      "id": "where-it-helps",
      "title": "Where It Helps"
    },
    {
      "id": "where-it-fails",
      "title": "Where It Fails"
    },
    {
      "id": "academic-connections",
      "title": "Academic Connections"
    },
    {
      "id": "practical-checklist",
      "title": "Practical Checklist"
    },
    {
      "id": "the-de-hype-check",
      "title": "The De-Hype Check"
    },
    {
      "id": "open-questions",
      "title": "Open Questions"
    }
  ]
}