Instructions file CodexOpenCode
AGENTS.md instructions for goemktg/Prompt_Template, covering agents.md, scope & references, project overview, repository structure and development workflow.
Instructions file CodexOpenCode
AGENTS.md instructions for goemktg/Prompt_Template, covering agents.md, scope & references, project overview, repository structure and development workflow.
Agent
System architecture design and technical solution planning. Designs component boundaries, validates architectural decisions, and coordinates with planners and generators. Triggers: design system, structure components, architecture decision, component boundary, where should X go.
Agent
Research lineage mapper via depth-first citation chaining. Identifies foundational papers and research evolution from a seed paper. Triggers: find foundational papers, citation chain, research lineage, trace citations, paper genealogy.
Agent
Production code generation with pre-delivery validation and execution verification. Generates working, tested code from specifications. Triggers: implement X, add feature Y, write function, generate code, create module. SRP: Code generation with self-validation only (no planning or design).
Agent
Tier 1/2 Quality Gate for code validation before delivery. Validates production-readiness with blocking/non-blocking verdicts. Triggers: review my code, is this good, check for issues, quality gate, code review. SRP: Quality validation only (no fixes).
Agent
Thin protocol-layer mediator between @orchestrator and an optional deep-think reasoning protocol. Normalizes intake, runs protocol-local phase sequencing when the protocol asset exists, and returns either a final payload or redirect metadata for council escalation.
Agent
Documentation Quality Review for clarity, accuracy, completeness, and consistency. Returns APPROVED, CONDITIONAL, or REJECTED with structured findings. Triggers: review this doc, check documentation, doc quality, is this doc accurate. SRP: Review only (no writing).
Agent
Professional documentation generation for README, API docs, guides, tutorials, and examples. Produces developer-first documentation with clear examples and multiple skill levels. Triggers: write docs, document this, create README, API documentation, write guide.
Agent
Thin protocol-layer mediator between @orchestrator and an optional dt-council reasoning protocol. Sequences council-local phases, normalizes the result, and keeps council execution separate from repository lifecycle ownership.
Agent
Execution specialist for shell-native commands and runtime tasks. Uses terminal execution for shell-native tools and routes Python execution through pylance-mcp-server/pylanceRunCodeSnippet instead of terminal Python by default.
Agent
Knowledge curator that extracts reusable patterns from project history, failures, experiments, and reviews. Triggers: lessons learned, extract patterns, what did we learn, curate knowledge, post-mortem insights.
Agent
Autonomous bug diagnosis and targeted fixing with execution verification. Reproduces issues, identifies root causes, applies minimal fixes, and verifies via testing. Triggers: fix this, error in, not working, debug, test failing, runtime error.
Agent
Creative idea generator with ethical and safety analysis focus. Specializes in user-centric innovation, divergent thinking, and UX implications. Triggers: brainstorm ideas, alternative approaches, user-centric ideas, ethical analysis, safety-aware innovation.
Agent
Feasibility-focused idea generator with quantitative analysis. Specializes in technical viability, resource optimization, and market sizing. Triggers: cost-effective solution, optimize resources, feasible approach, budget-friendly ideas.
Agent
System architecture idea generator. Specializes in system design, business models, strategic scalability, and structural patterns. Triggers: system design ideas, scalable architecture, business model options, strategic approach.
Agent
Research-grounded prompt engineer: designs, creates, and edits prompt assets (agent definitions, skill files, prompt templates) grounded in the local paper database with technique citations. Triggers: write prompt, prompt planning, select technique, prompt blueprint, update catalog.
Agent
Mathematical correctness reviewer. Verifies equations, dimensional consistency, numerical stability, and mathematical properties of implementations against specifications. Triggers: verify formula, check math, is this equation correct, dimensional analysis, numerical stability.
Agent
Lifecycle-first orchestration planner for complex multi-agent workflows. Coordinates the INIT -> ATOMIZE -> PLAN -> EXECUTE -> REPORT -> AWAIT -> FINALIZE cycle, defines specialist sequence, and keeps implementation delegated. Triggers: plan this, multi-phase task, coordinate agents, complex workflow, decompose task.
Agent
Risk-aware planning with QA gates, safety verification, testing protocols, and pre-mortem analysis. Triggers: risk assessment, what could go wrong, safety analysis, test strategy, QA plan.
Agent
Feasibility and resource planning agent. Validates hard constraints, computes explicit budgets (memory, compute, bandwidth, time), and produces mathematically verified plans. Triggers: is this feasible, estimate resources, budget planning, resource constraints.
Agent
Architecture-first planning agent. Produces component blueprints, interface contracts, Mermaid diagrams, and structural implementation strategies. Triggers: plan the architecture, design strategy, component blueprint, interface design.
Agent
Execution-based quality verification. Runs tests, detects regressions, identifies flaky tests, and generates reproduction scripts. Triggers: run tests, check for regressions, flaky test, test reliability, regression check.
Agent
System complexity, safety, and resource analysis specialist. Performs deep research on constraints, failure modes, and engineering trade-offs. Triggers: system constraints, failure analysis, complexity research, resource limits, safety research.
Agent
Implementation-focused research agent. Analyzes papers for practical methods, code patterns, computational requirements, and reproducible setups. Triggers: how to implement X, API usage for, implementation guide, practical methods, code patterns.