6.8 KiB
6.8 KiB
Intent of this file: This file is intended for new-task-prompt development
and orchestrator templates. It is NOT for actual task research and creation. If
your current job is to create, plan, or format a task file, please refer to
tasks/GUIDELINES.md and ignore the templates in this file.
Orchestrator Meta-Prompts (tasks/path.md)
1. Task Planning Template (System Analyst)
Use this template to instruct an agent (such as Jules) to take a technical
idea, critically analyze the existing codebase, identify risks/alternatives, and
draft a compliant task file in tasks/new/.
**Role:** Act as a System Analyst and Solutions Architect. Do not execute code or modify the codebase during this session. Your sole deliverable is to generate a task file in tasks/new/.
**The Objective:** [State exactly what you want to achieve.]
**Suggestions & Initial Thoughts:**
*Here is our starting hypothesis. Treat these as ideas to evaluate, not rigid requirements:*
* [Suggestion 1]
* [Suggestion 2]
**Your Task:**
1. Research the codebase and analyze this objective against our current architecture.
2. Critically evaluate our suggestions: Do they make technical sense, or is there a cleaner, more native, or more modular approach?
3. Map out the `Target Files`, prerequisites/dependencies, and explicitly identify potential architectural risks, regressions, or friction.
4. Draft the task file in `tasks/new/` strictly adhering to the naming conventions, header metadata, and mandatory structure defined in `tasks/GUIDELINES.md`.
2. Implementation Execution Template (Developer)
Use this template when ready to command an agent to execute the work outlined in an existing task file.
**Role:** Act as an Implementation Engineer.
**The Task:** Please review and execute the plan detailed in `[tasks/new/YYYY-MMDD.XX.agent.type.scope.title-HHMM.md]`.
**Directives:**
1. Follow the strict 4-step state machine (Research -> Implementation -> Quality Gates -> Review).
2. Apply minimal, pure functional modifications.
3. Ensure all downstream pipeline steps and context contracts remain intact.
4. Run all quality gates (`deno fmt`, `deno task lint`, `deno task check`, `deno task test`).
5. Run ONLY unit tests (`deno task test` / `deno test --allow-all`). Do NOT execute `deno task start` or `deno task dev` (live database daemons are not running in sandbox containers).
6. Upon successful completion and verification, update the task status according to `tasks/GUIDELINES.md`.
3. Implementation & Code Quality Audit Template (tasks/audit-2.md)
Use this template to perform an exhaustive post-implementation audit on code
changes, PR branches, or completed tasks before merging or deploying. See
complete protocol in
tasks/audit-2.md.
**Role:** Act as a Senior Security, Architecture & Quality Assurance Auditor. Do not modify production code or push commits during this audit session. Your sole deliverable is an objective Code Quality & Security Audit Report.
**The Scope:** [Specify target branch, recent commits, modified files, or completed task in `tasks/wip/TASK_FILENAME.md`].
**Your Task:**
Audit the actual codebase changes against the original task specification and our system guidelines (`AGENTS.md`, `docs/`) using the 5-point implementation rubric in `tasks/audit-2.md`:
1. Specification & Scope Fidelity
2. Zero-Trust Security & Boundary Safety
3. Modularity & Clean Architecture
4. Quality Gates & Test Evidence
5. Runtime & Client-Side Integrity
**Deliverable:**
Provide a structured code audit report with:
- **Overall Implementation Score:** [e.g. 5/5, 4/5, or Letter Grade]
- **Verdict:** [MERGE READY / REVISION REQUIRED / BLOCKED]
- **Successes & Verified Achievements:** Exact functionality verified and working as expected.
- **Deficiencies, Regressions & Failure Risks:** Technical bugs, unhandled errors, UX regressions, or security gaps found.
- **Actionable Remediation Suggestions:** Specific, prioritized code fixes or refactoring steps needed before final approval.
4. Task Plan Review & Critique Template (tasks/audit-1.md)
Use this template to critically audit, score, and refine new task files in
tasks/new/ before approving them for implementation. See complete protocol in
tasks/audit-1.md.
**Role:** Act as a Principal Systems Architect and Task Quality Critic. Do not modify the codebase or execute code during this session. Your sole deliverable is a structured task critique report.
**The Scope:** Review the newly generated task file in `[tasks/new/TASK_FILENAME.md]` against our master architectural guidelines (`AGENTS.md`, `docs/`, `tasks/GUIDELINES.md`).
**Your Task:**
Critically evaluate the proposed task plan against the 5-point quality rubric in `tasks/audit-1.md`:
1. Metadata & Standards Compliance
2. Architectural Alignment & Safety
3. Purity & Simplicity (Avoiding bloat)
4. Implementation Precision & State Machine
5. Testing & Quality Gate Rigor
**Deliverable:**
Provide a structured critique report with:
- **Overall Score / Rating:** [e.g. 5/5, 4/5, or Letter Grade]
- **Verdict:** [APPROVED / NEEDS REVISION / REJECTED]
- **Strengths & Plan Alignment:** Key architectural insights captured in the plan.
- **Identified Gaps & Potential Risks:** Missing edge cases, unaddressed failure modes, or UX/architectural oversights.
- **Concrete Suggested Revisions:** Exact adjustments, additions, or rewordings to incorporate into the task file before implementation starts.
5. Architecture & Root-Cause Investigation Template (Investigator)
Use this template to instruct an agent to deeply investigate complex bugs, reproduction anomalies, or distributed edge cases without modifying production code.
**Role:** Act as a Principal Systems & Security Investigator.
**The Scope:** [Describe anomaly, bug symptoms, error logs, and affected components.]
**Directives:**
- **STRICT CONSTRAINT:** Experiment, research, and report ONLY. Do not modify existing production code or create PRs with production changes.
- Place all reproduction test scripts and experiments in `scratch/` or hermetic test harnesses.
**Your Task:**
1. Formulate clear, falsifiable hypotheses based on observed logs and symptoms.
2. Design and execute minimal reproduction scripts or benchmarks to test each hypothesis in isolation.
3. Trace data flows through database, caching layers, ingress proxies, and client runtime engines.
4. Document the definitive root cause and rank potential architectural solutions from most reliable to least, with explicit rationale and trade-offs.
**Deliverable:**
Author an investigative report in `scratch/INVESTIGATIVE_REPORT.md` detailing verified findings, discarded hypotheses, and ranked solutions.