{"id":"b7ce2c2e-8d06-42ec-96a1-204e52d89917","slug":"problem-solving-systems-thinking","kind":"skill","name":"problem-solving-systems-thinking","summary":"This skill helps an agent decompose complex problems, map workflows, identify decision points, and design an agent's end-to-end flow from intake through validation and final response.","owner":"bybaochaule","repoUrl":"https://github.com/bybaochaule/problem-solving-systems-thinking","license":"MIT","licenseTier":"permissive","trust":"community","stars":3,"match":null,"fewStars":true,"format":"skill","createdAt":"2026-06-22T22:17:38.000Z","addedAt":"2026-10-21T11:08:04.044Z","logoUrl":"https://avatars.githubusercontent.com/u/293674774?u=75204f8934d2ae48401684340cbb25d15e3f4176&v=4","linkOnly":false,"collection":null,"skillPath":null,"categories":["design-media"],"tags":["chatgpt-skills","problem-solving","problem-solving-skills","systems-thinking"],"runtime":"binary","local":true,"cloudCapable":false,"cloudCapability":"unknown","cloudReadyOnPlatform":false,"needsApiKey":false,"reviewed":false,"syncedAt":"2026-10-22T12:43:01.398Z","sample":false,"url":"https://conduits.dev/tools/problem-solving-systems-thinking","language":"Python","transports":["stdio"],"install":{},"requirements":{"env":[],"upstreamMayCharge":false},"tools":null,"readme":"# Problem Solving & Systems Thinking\n\n## Overview\n\nThis skill helps an agent decompose complex problems, map workflows, identify decision points, and design an agent's end-to-end flow from intake through validation and final response.\n\nUse it when a user asks for process design, systems thinking, workflow mapping, lead generation flow design, decision trees, SOP design, bottleneck analysis, or agent behavior design.\n\n## What it creates\n\nThe skill can produce:\n\n- Problem decompositions and system boundaries.\n- Current-state and future-state workflow maps.\n- Decision tables with rules, branches, thresholds, and escalation paths.\n- Agent flow specs covering how the agent thinks, decides, acts, validates, and escalates.\n- Mermaid diagrams, SOPs, state machines, RACI-style handoff tables, and implementation-ready process specs.\n- Risk, bottleneck, and missing-information analysis.\n\n## Example user prompts\n\n- \"Map our lead generation process from prospecting to sales handoff.\"\n- \"Decompose this messy customer onboarding problem and design the workflow.\"\n- \"Create a decision tree for support ticket triage.\"\n- \"Design how an agent should think, decide, and act for invoice review.\"\n- \"Turn this process into a state machine with escalation paths.\"\n\n## Contents\n\n- `SKILL.md`: agent-facing instructions, trigger description, boundaries, workflow, decision policy, output formats, quality checks, and safety rules.\n- `README.md`: human-readable overview and usage notes.\n- `agents/openai.yaml`: OpenAI metadata, default prompt, invocation policy, and file dependencies.\n- `references/style-guide.md`: style rules, output patterns, terminology, edge cases, and anti-patterns.\n- `references/lead-generation-example.md`: worked example for mapping a lead generation workflow.\n- `assets/workflow-map-template.md`: reusable workflow mapping template.\n- `scripts/example_helper.py`: deterministic helper that checks whether a Markdown workflow map includes key sections.\n- `assets/.gitkeep`: keeps the assets directory in version control.\n\n## Package structure\n\n```text\nproblem-solving-systems-thinking/\n|-- SKILL.md\n|-- README.md\n|-- agents/\n|   `-- openai.yaml\n|-- assets/\n|   |-- .gitkeep\n|   `-- workflow-map-template.md\n|-- references/\n|   |-- style-guide.md\n|   `-- lead-generation-example.md\n`-- scripts/\n    `-- example_helper.py\n```\n\n## Suggested default output\n\n```markdown\n# [Process or Problem Name]\n\n## 1. Goal and success metric\n## 2. System boundary and assumptions\n## 3. Actors, tools, inputs, and outputs\n## 4. Workflow map\n## 5. Decision points\n## 6. Agent flow\n## 7. Risks, bottlenecks, and mitigations\n## 8. Suggested next steps\n```\n\n## Safety notes\n\nThe skill is for analysis, design, and planning. It should not execute workflows, contact people, scrape data, update systems, or make high-stakes determinations without explicit authorization, safe tooling, and human review where appropriate.\n"}